changes in the delivery of ecosystem services in galveston ... · changes in the delivery of...
TRANSCRIPT
1
Changes in the Delivery of Ecosystem Services in Galveston Bay, TX, under a Sea-Level Rise Scenario
G. Guannel, A. Guerry, J. Brenner, J. Faries, M. Thompson, J. Silver, R. Griffin, J. Proft, M. Carey, J. Toft, G. Verutes
November 2014
2
Table of Contents Abstract ......................................................................................................................................................... 3 1 Introduction ........................................................................................................................................... 4 2 Physical Setting ..................................................................................................................................... 7
2.1 Geophysical setting ....................................................................................................................... 7 2.2 Climate Stressors .......................................................................................................................... 9 2.3 Ecosystem Services in the Bay and Impacts of Climate Change ................................................ 12
3 Ecosystem Service Quantification Methods ....................................................................................... 14 3.1 Modeling Scenarios .................................................................................................................... 14 3.2 Coastal Protection ....................................................................................................................... 15
3.2.1 Storm water Retention ........................................................................................................ 16
3.2.2 Storm Surge ........................................................................................................................ 16
3.2.3 Nearshore Waves ................................................................................................................ 17
3.2.4 Valuation ............................................................................................................................. 23
3.2.5 Structures ............................................................................................................................ 27
3.2.6 Forcing Conditions .............................................................................................................. 27
3.3 Blue Carbon ................................................................................................................................ 30 3.3.1 Carbon Storage and Sequestration ...................................................................................... 30
3.3.2 Valuation ............................................................................................................................. 31
3.4 Fisheries ...................................................................................................................................... 32 3.4.1 Fishery Habitat Change ....................................................................................................... 32
3.4.2 Shrimp Fishery Model ........................................................................................................ 34
4 Quantification of Ecosystem Services in the Bay ............................................................................... 36 4.1 Coastal Protection ....................................................................................................................... 36
4.1.1 Wetlands as Storm Buffer and Storm Water Retention Capacity ....................................... 36
4.1.2 Storm Impact Reduction and Avoided Damages due to Wetlands ..................................... 37
4.1.3 Structures ............................................................................................................................ 45
4.2 Blue Carbon ................................................................................................................................ 47 4.3 Fisheries ...................................................................................................................................... 56
4.3.1 Fisheries Habitat Change .................................................................................................... 57
4.3.2 Shrimp Population Growth ................................................................................................. 58
5 Protecting Galveston Bay Wetlands .................................................................................................. 59 6 Conclusion .......................................................................................................................................... 63 7 Future Work ........................................................................................................................................ 66 8 References ......................................................................................................................................... 116
3
Abstract
In this document, we quantify the coastal protection, fisheries and carbon storage and
sequestration services delivered by wetlands in Galveston Bay, currently and under a 1 m sea-
level rise through 2100. During coastal storms, wetlands have the capacity to store millions of
cubic meters of storm water and attenuate total water levels. They also can avoid millions of
dollars in property damage and protect hundreds of people against the impact of storms. Those
services can continue to be delivered as sea-level rises, as long as wetlands are allowed to
migrate. In addition, wetlands currently avoid millions of dollars of environmental damages by
sequestering carbon. However, if sea level rises 1 m by 2100, the amount of CO2 released in the
atmosphere by defunct wetlands might cause millions of dollars of environmental damages.
Finally, it appears that most fisheries species might benefit, in the form of increased populations,
from more available wetland habitats as sea level rises.
This report is a summary of the work conducted under NOAA’s Sectoral Applications
Research Program (SARP) Grant Number NA11OAR4310136.
4
1 Introduction
Galveston Bay, TX (the Bay) and its surroundings are home to more than half a million
people, excluding the greater Houston area (USCB 2013). The low-lying areas surrounding the
Bay include more than 1,000 km2 of terrestrial, freshwater and coastal wetlands (WPC 2011),
which provide a suite of diverse and valuable services, or benefits, to communities in the Gulf of
Mexico. Wetlands have the potential to reduce storm surges, wave height, and the height of
structures required to protect people and property from coastal hazards (Wamsley et al. 2010;
Shepard et al. 2011; The Bay Institute 2013). Wetlands can also maintain and improve the health
of key fisheries (Zimmerman et al. 2002; Minello et al. 2008; Bohannon 2013), provide habitat
for valued bird species (Lester & Gonzalez 2011; U.S. Fish and Wildlife Service 1997), and
provide recreational opportunities (Lester & Gonzalez 2011; Yoskowitz et al. 2012).
Additionally, they can store and sequester important quantities of carbon (Siikamäki et al. 2013;
Feagin et al. 2010).
Global climate change impacts, in the form of increases in air and water temperatures,
and ocean acidity (IPCC 2013) have the potential to negatively affect coastal wetlands and the
ecosystem services they deliver (Burkett & Kusler 2000; Erwin 2008; Millennium Ecosystem
Assessment 2005). However, one of the most pressing and visible threats to the bay and its
residents is the rise in sea level that it experiences (Rice 2013; Yoskowitz et al. 2009). Galveston
Bay experienced substantial land loss due to coastal development, shoreline erosion, land
subsidence and sea-level rise (SLR) (Yoskowitz et al. 2009). In turn, these shoreline changes
make the region even more vulnerable to sea-level rise, which might lead to higher storm-surge-
levels and inland inundation levels (Leatherman 1984; Warner & Tissot 2012). Impacts from
these stressors could result in the displacement of more than 100,000 households and create more
5
than $12 billion in infrastructure losses (Yoskowitz et al. 2009). Consequently, it is likely that
shoreline protection measures will be implemented in order to limit the impacts of coastal
hazards on communities.
Typical coastal protective measures have been the construction of seawalls and other
structural/engineered solutions. For instance, the 17 feet high, 10 miles long Galveston Seawall,
built in response to the deadly Hurricane of 1900, was once one of the longest in the world.
While such structures can provide protection against storm inundation they can be costly and
require constant maintenance and upgrade (USACE 2002), especially in the face of a changing
climate (Sekimoto et al. 2013). Shore-parallel structures also have adverse impacts on nearshore
environments. In Galveston, they increase the erosion of adjacent shoreline (Leatherman 1984;
Wallace et al. 2009) and fix the shoreline in place, which can yield deeper water depths
nearshore and the disappearance of intertidal habitats (Dugan & Hubbard 2009; USACE 2002;
Miles et al. 2001). In addition, by creating a barrier between marsh areas and their surrounding
environment, shoreline structures can lead to the modification and/or disappearance of certain
types of coastal wetlands that depend on regular or irregular exposure to salt or brackish waters
(Bozek & Burdick 2005; Mattheus et al. 2010). These losses can negatively impact the amount
and quality of ecosystem services provided by these important habitats.
Ecosystem services, and their continuous delivery, are essential for the health and long-
term sustainability of communities. However, as mentioned above, communities sometimes put
in place protective measures that alter the delivery of services to coastal communities.
Consequently, it is important to identify and quantify the different services delivered by natural
habitats in order to properly account for their loss when those habitats are removed or degraded
directly by humans or indirectly by climate change impacts.
6
In this study, we quantify the changes in the delivery of three ecosystem services –
coastal protection, carbon storage and sequestration, and fisheries – in Galveston Bay, TX, under
a 1 m by 2100 sea-level rise scenario. We also estimate the amount of avoided damages and the
number of people protected by wetlands against the impacts of storms under current and future
environmental conditions. We compute the amount and value of carbon stored, sequestered and
re-emitted in the Bay. Finally, we explore changes in the amount of habitat for a number of
important fisheries in the Bay and model expected changes in brown shrimp landings.
The report is organized as follows. In Section 2, we present the physical setting of Galveston
Bay, discuss climate stressors, and introduce some of the ecosystem services the Bay’s wetlands
deliver to people. In Section 3 we present the scenarios we chose to explore and the methods we
used to quantify changes in the delivery of coastal benefits under those scenarios. In Section 4
we present the results of our analyses. In Section 5 we introduce the protection and restoration of
oyster reefs as a mechanism for minimizing the adverse impacts of sea level rise on wetlands.
We present conclusions and recommendations in Section 6.
7
2 Physical Setting
2.1 Geophysical setting
Galveston Bay (the Bay) is sheltered from the Gulf of Mexico by Galveston Island and
the Bolivar Peninsula. It is the 7th largest estuary in the U.S., and home to the Port of Houston,
the 2nd busiest port in the U.S. in terms of tonnage (Port of Houston Authority 2012). This 1,600
km2 wide, shallow, micro-tidal estuary has an approximate tidal range of 0.62 m [MHHW is 0.28
m above MSL (NOAA-CSC, 2013)] and an average depth of 2.4 m. The Bay can be separated
into 4 distinct regions: West Bay, East Bay, Trinity Bay and Upper Galveston Bay (Figure 1).
Twenty percent of the Bay’s shores are protected by man-made structures, such as
seawalls or riprap revetments (NOAA, 2012). The other major shoreline types are gravel and
sandy beaches (10%), muddy shore banks (15%) and wetlands (52%) (Figure 2). Wetlands,
oyster reefs, and seagrasses are the primary living coastal habitats in the Bay. Seagrasses
currently cover only 0.1% of the Bay’s bed (1.7 out of 1,550 km2). Although the amount of
seagrass in the Bay has increased slightly in the recent past, it is unlikely that gains will be
sufficient enough to match its historical coverage of more than 10 km2 (Dahl 2011; Lester &
Gonzalez 2011). Thus, we focus our analyses on oyster reefs and wetlands.
8
Figure 1: Living Habitats in Galveston Bay
Most of the Bay’s oyster reefs are subtidal and located in the middle of the Bay, more
than 1 km from the shore. Reefs occupy approximately 6.5% of the Bay’s bed (TPWD 2010),
and this number is slowly increasing thanks to recent restoration efforts by the Texas Parks and
Wildlife Department and others. Oyster reefs are typically severely impacted by hurricanes: over
60% of the oyster reefs in Galveston Bay were destroyed during Hurricane Ike, and despite
extensive ongoing restoration efforts, reefs have still not recovered (Haby et al. 2009). We
explore the role of oyster reefs in protecting marshes in Section 5.
Estuarine wetlands cover approximately 570 km2 of the land in and around Galveston
Bay (Figure 1), roughly half of the Bay’s water surface area. These wetlands are primarily
composed of salt marsh and brackish marsh communities (Appendix A), which are common
habitats for many fisheries species in the Bay. Although the current acreage of marsh has
declined as compared to coverage in the 1950’s, groundwater pumping regulations and habitat
9
conservation and restoration efforts have slowed, if not halted, the rate of salt marsh loss (Lester
& Gonzalez 2011). Given that development is forecasted to continue to occur within or directly
adjacent to existing development (H-GAC 2013), most of the future marsh losses are likely to be
fragmented marshes near current population centers (Appendix B).
Figure 2: Shoreline classification in Galveston Bay [data from NOAA (2012)].
2.2 Climate Stressors
Positioned along the Gulf Coast, the greater Galveston Bay region is regularly impacted
by tropical storms and hurricanes. Since the 1860’s, 32 hurricanes or tropical storms have made
landfall or passed within a 30 mile radius of Galveston Bay (NOAA 2014a). Each of these
storms generates high waves and surge in the bay. The 25 year return surge height (surge that has
a 4% chance of being exceeded in any one year) is estimated to be approximately 3 m, and the
100 year return surge height is estimated to be above 5 m (Keim et al. 2007; Needham & Keim
2011; Figure 3). In particular, Hurricane Ike made landfall on the north end of Galveston Island
10
in 2008 as a Category 2 hurricane and became the most expensive hurricane in Texas history,
causing an estimated 29.5 billion dollars in property damage (Blake et al. 2011).
Figure 3: Storm surge return period in Galveston Bay
Increases in the frequency and magnitude of storms, air and sea-surface temperature, sea
water salinity and sea-level are all likely to impact the Bay (Yoskowitz et al. 2009; Elsner et al.
2008; Leatherman 1984; Texas Seagrant 2013) and to intensify the impacts of storms (Warner &
Tissot 2012). In this work, we focus on the impacts of sea-level rise, one component of climatic
change that has already been observed to occur in Galveston Bay and that is likely to have wide-
ranging impacts (Yoskowitz et al. 2009; Hallegatte et al. 2013; Texas Seagrant 2013;
Leatherman 1984). (Given the concerns about vulnerability to coastal hazards and the possibility
that increases in storm intensity and frequency could exacerbate the effects of increases in sea
level, we analyzed 33 years of hindcast wind and wave data just offshore of the Bay (WIS
dataset (USACE 2010) and NOAA buoy 42035 (NDBC 2014). We did not measure an increase
in observed wave height or wind speed in the region over that time period.) Sea-level is
estimated to be rising in the Bay at approximately 6.84 mm/yr (NOAA 2014b). This rise will
cause wetlands to migrate landwards and cause major shifts in the amount and distributions of
11
natural habitats in and around the Bay. This rate of sea-level rise also includes local subsidence,
which is higher than many other areas in the Gulf.
Warren Pinnacle Consulting ( 2011) simulated the changes in land class types around the
Bay under different sea-level rise scenarios using a model named Sea-Level Affecting Marshes
Model (SLAMM, WPC 2009; Park et al. 1986). Results for the 1 m sea-level rise by 2100
scenario, -- a reasonable estimate since sea level is expected to increase by approximately 70 cm
over 100 years (NOAA 2014b) -- predict a net loss of marine and freshwater wetlands and a net
gain of open water areas (Figure 4, Figure 5).
Figure 4: Marine and fresh marsh coverage around the Bay in 2004, 2050 and 2100 under a 1 m SLR scenario.
12
Figure 5: Change in area of different habitat classes in 2050 and 2100, as predicted by SLAMM, for the 1 m SLR scenario.
2.3 Ecosystem Services in the Bay and Impacts of Climate Change
Healthy natural habitats provide valuable and essential services to humans (Millennium
Ecosystem Assessment 2005; Tallis et al. 2008). In Galveston Bay, seagrasses, oyster reefs and
wetlands protect coastlines from coastal hazards, support fisheries, and provide carbon storage
and sequestration services, among other benefits. Table 1 describes some of those services
(Yoskowitz et al. 2012; Lester & Gonzalez 2011); italicized text represents the services that are
modeled herein. Among these three habitats, coastal wetlands represent the land cover type that
will be predominantly affected by sea-level rise stressors since they are located at the water’s
edge.
Wetlands have the potential to protect people and property against the impact of coastal
hazards. They act as a buffer against the impacts of storms and chronic erosion because their
very existence can prevent people from building and living directly along the shoreline. They
also attenuate storm surge and waves, thus reducing damages during storm events (Anderson et
13
al. 2011; Jadhav et al. 2013; Wamsley et al. 2010). Finally, they can reduce the size of
engineered protective structures, including levees, or dikes (The Bay Institute 2013).
Additionally, wetlands are natural stores of carbon, which accumulates in the plant biomass and
the soil beneath plants. Lastly, wetlands support the growth of the major commercial and
recreational fisheries in the Bay (Bohannon 2013; Orth & van Montfrans 1990; Turner 1977;
Zimmerman et al. 2002), which include white and brown shrimp, blue crab, southern flounder
and red drum. The delivery of these services can be altered by sea-level rise if marshes are not
allowed to naturally migrate landward, which can lead to a reduction of their footprint and the
services they provide.
Table 1: Habitats in Galveston Bay and the Some of the Key Services they Provide
Habitat Service Wetlands Seagrass meadows Oyster Reefs
Coastal Protection
Wave and surge attenuation; storm water retention; sediments stabilization
Wave attenuation; sediments
stabilization Wave attenuation
Fisheries Habitats for certain species Habitats for certain species
Habitat and source of food for certain
species
Blue Carbon Storage and sequestration Storage and sequestration
Storage and sequestration
14
3 Ecosystem Service Quantification Methods
3.1 Modeling Scenarios
We quantify the coastal protection, blue carbon and fisheries services delivered by
wetlands in Galveston Bay currently, and under a future scenario where sea level increases 1 m
by 2100, from its 1992 baseline. This scenario, which also predicts nearly 0.7 m of sea-level rise
by 2050 (WPC 2011), captures the current trend of sea-level rise in the Bay (NOAA 2014b).
The primary inputs used in our analysis are land cover class distribution maps around the
Bay for 2004 (current), as well as SLAMM modeled projections for 2050 and 2100. The initial
land cover maps were generated primarily by converting the National Wetlands Inventory
(USFWS 2009) maps to categories used in SLAMM. Future land cover projections come from
previous efforts using SLAMM in Galveston Bay (WPC 2011) (Figure 4). SLAMM converts
different land cover classes, and wetland types in particular, into other classes based on the rate
of sea-level rise, nearshore topography, local land subsidence, shoreline erosion and wetland
accretion, among other parameters. In particular, it converts land to new wetland habitat as sea-
level rises, unless barriers to wetland migration exist. This assumption in the model can lead to
unrealistic results since local communities might not allow natural marsh migration to occur in
certain regions that currently do not contain any physical barriers.
In order to reflect the extent of that model’s limitations, we also consider, in the
quantification of coastal protection and blue carbon services, a scenario where no new wetlands
areas are created. We did not include such a scenario in the fisheries model because it didn’t
yield any significant difference in outputs. To create land use-land cover maps for this scenario,
we replaced by estuarine water all regularly and irregularly flooded salt marsh wetlands
predicted to migrate in 2050 and 2100 onto regions that do not contain wetlands under the initial
15
conditions. These maps might over-estimate the amount of estuarine habitats present in the future
as local communities might put into place different mitigation measures to prevent land from
being inundated.
The potential future land cover maps described above are used to quantify the changes in
the delivery of ecosystem services delivered by wetlands as sea level rises. The models and
scenario parameters for each service are described in further details in the remainder of this
document.
3.2 Coastal Protection
The coastal protection services delivered by wetlands during storms under current and
future conditions (Section 3.2.6) are quantified using four different metrics: 1) the storm water
retention capacity of the wetlands, 2) the amount of avoided increase in total water level around
the Bay due to the presence of wetlands, which is translated into 3) the amount of avoided
damages, and 4) the reduction in required levee height needed for protection due to the presence
of the marsh, in the West Bay region. We perform the last two analyses in the West Bay region
because it is where we had access to parcel data and the U.S. Federal Emergency Management
Agency (FEMA) FIRM database (see Section 3.2.4).
The total water level during storms is defined as the sum of storm surge, wave height and
wave-induced setup. It is computed by modeling storm surge and wave evolution over a series of
transects drawn perpendicular to the coast, starting a few meters offshore of the 2004 shore
boundary (Figure 6). While this approach neglects 2-D hydrodynamic processes, it is
conceptually equivalent to the process and standards of the FEMA approach (FEMA & FIS
1988).
16
Figure 6: Transect layout around Galveston Bay. Black lines represent modeling transects. Red line indicates the hypothetical levee location.
3.2.1 Storm Water Retention
Wetlands are highly complex three dimensional features that can retain water because of
the permeability of their soil as well as their topography. We estimate the storm water retention
capacity of wetlands by computing the total volume of water that they can retain if they were to
become fully submerged.
3.2.2 Storm Surge
We model storm surge overland in the presence of wetlands by taking a modified bathtub
approach, where coastal land is progressively inundated until the local topography exceeds the
surge elevation. For each 1-D transect used in the analysis (Figure 6), we project the surge
landward from the offshore boundary of that transect. When surge waters encounter wetlands,
we reduce the water elevation using a fixed attenuation rate (Table 2). This approach captures the
17
uncertainty in attenuation rate values associated with storm characteristics and path, topography
and configuration of the wetland captured in Figure 7 (see Appendix C). It also illustrates that
the potential for wetlands to reduce storm surge is inversely proportional to the storm strength
(Appendix C).
Table 2: Default surge attenuation rates by wetlands
Surge at Coast [m] 0.5 1 1.5 2 2.5 3 3.5 4 Attenuation Rate [cm/km] 20 17.4 14.8 12.2 9.6 7 4.5 2
Figure 7: Rates of surge reduction as measured by different studies. In addition to rates measured by other researchers, we also quantified the potential for marshes in the Bay to reduce the surge generated by 3 idealized storms in West Bay (see Appendix C).
3.2.3 Nearshore Waves
In addition to reducing storm surge, wetlands can also reduce wave height and wave-
induced mean water level during storms. From values of significant wave height and peak period
18
at the seaward boundary of a particular transect, the model computes wave height and wave
setup at each point along that transect following Equation (1):
(1)
where is the wave energy density, is the wave group velocity. In Equation (1), waves are
re-generated by winds via the source term (FEMA & FIS, 1988), which is a function of the
sustained wind speed of the storm, wave height and the local water depth ℎ. The water depth is
equal to the sum of the still-water depth – zero for locations on land – and the surge height.
Waves also dissipate their energy via breaking, in Equation (1) and bottom friction,
. Those terms are expressed following the standard formulation of Thornton & Guza (1983).
The expression of 𝐷𝑓 is a function of the bed friction coefficient 𝐶𝑓, which is fixed at
for sandy beds, among other terms. 𝐶𝑓 for agricultural and bare lands is computed from their
corresponding Manning coefficients 𝑀𝑛, following the classic conversion expression (Augustin
et al. 2009):
𝐶𝑓 =8𝑔ℎ1 3⁄ 𝑀𝑛 (2)
w gWind b f v
E CS D D D
x∂
= − − −∂
wE gC
WindS
bD
fD
0.01fC =
19
Finally, wave energy is dissipated because of the presence of vegetation via the term ,
following the expression in Mendez & Losada (2004). It is a function of wetlands density, 𝑁𝑣,
stem diameter, 𝑑𝑣, and a drag coefficient, 𝐶𝑑, which is approximately equal to 0.1 (FEMA
2007b).
In the absence of detailed local field measurements, we created a database of average
plants’ physical characteristics in Galveston Bay (
vD
20
Table 3; see also Appendix A). Based on observations by Feagin et al. (2011), this
database was modified to create “healthy” and “degraded” physical parameter values for each
habitat. For the healthy habitat conditions, average plant densities presented in
21
Table 3 were increased by 50%, and average plant diameter and height by 30%. For
degraded habitat conditions, average plant densities were reduced by 50%, and average plant
diameter and height by 30%. These parameter values are not intended to accurately reflect the
status of, and expected changes to, each habitat. Rather they are used to evaluate the model’s
sensitivity to changes in these input parameters. Outputs provide an uncertainty estimate of their
protective capacity given changes in their status. (Guannel et al. (2014) performed an uncertainty
analysis of Equation (1) and showed the importance of the choice of drag coefficient in
estimating the role of habitats in reducing the impacts of wave height.)
22
Table 3: Physical Characteristics of Vegetation Associated With Different Land Classes
Land Cover Class Height [m]
Diameter [m]
Density [units/m2] Source
Swamps and Forested Wetlands 15 1.5 0.0172 Arborday.org
Scrub/Shrub 2.5 0.1016 0.0861 University of Florida, Alabama Forestry Department, and Davesgarden.com
Forests 10 0.2 0.0861 FEMA
Low Salt Marsh 0.8 6.4x10-3 172 FEMA guidance for the Chenier Plain Region
High Salt Marsh/Brackish Marsh
0.57 5.8x10-4 3,585 FEMA guidance for the Chenier Plain Region
Fresh Marsh 1.22 8.14x10-3 107 FEMA guidance for the Chenier Plain Region
Equation (1) applies to vegetated and non-vegetated bare beds. When waves encounter
structures or obstacles such as roads or buildings, isolated or in a residential community, we
compute the wave height, , landward of those wave-blocks following (FEMA & FIS, 1988):
𝐻𝑡 = max � 𝛾ℎ𝑡𝐾𝑡𝐻𝑖 (3)
where 𝛾 is a breaking coefficient used in the breaking dissipation expression 𝐷𝑏 and ℎ𝑡 is the
water depth landward of the blocks. The variable 𝐻𝑖 is the wave height seaward of a structure
block, which has a transmission coefficient 𝐾𝑡. We compute 𝐾𝑡 as (FEMA & FIS 1988):
𝐾𝑡 = 𝐿𝑆𝑁𝑅 2⁄ (4)
tH
23
where 𝐿𝑆 is the length of a structure block and 𝑁𝑅 is the number of structures in that block.
From computed profiles of wave height, we estimate wave-induced mean water level
following (Guannel et al. 2014):
(5)
where 𝜂 is the mean water level, ℎ is the water depth, 𝑆𝑥𝑥 is the wave-induced radiation stress, 𝐻
the wave height, and 𝑘 the wave number.
3.2.4 Valuation
The value of wetlands is defined herein as the dollar amount of avoided damages to
buildings (residential, commercial and industrial) and the number of people protected by those
habitats during a storm. The amount of avoided damages is estimated by taking the difference
between modeled property damages caused by storms in the presence and absence of wetlands
(where in the case of “absence” wetlands are replaced by bare land). The number of people
protected by wetlands is computed by taking the difference between the number of people
affected by the storm in the presence and absence of wetlands. Because of data limitations,
estimates of damages and number of people affected by storms are computed in the West Bay
region only (Figure 9).
The total dollar amount of damages 𝐷 caused by a storm is estimated as
( ) 33 0tanh16
xxd v v
S gkg h C d N Hx x khηρ η ρ
π∂∂
+ + − =∂ ∂
24
𝐷 = (𝑠 + 𝑧)𝑉 (6)
where 𝑉 is the value of a building (commercial or residential), and 𝑠 and 𝑧 are the structural and
content damage ratios (0 ≤ {𝑠, 𝑧} ≤ 1). We obtained a spatial inventory of the properties in that
region from the Galveston County appraisal district (GCAD 2014). Damage ratios values are
then computed as a function of the flood depth in a building following depth-damage
relationships used in FEMA’s HAZUS model (Figure 9) (Scawthorn et al. 2006; FEMA 2011;
Cannon et al. 1995; USACE 2003). Flood depth is defined as the total water level (Sections 3.2.2
and 3.2.3) in a building, relative to the first floor elevation of that building.
Because of a lack of detailed information about the location and physical characteristics
of buildings in West-Bay, various simplifications were made. First, we approximated the
location of structures at the centroid of the parcel where they were listed. Also, instead of relying
on a replacement cost approach used by HAZUS, we used parcel tax assessment data as a proxy
for a structure’s value (Cannon et al. 1995). Next, since we did not have parcel-level information
on the number of floors for residential buildings, we created a composite damage function for
single story and multi-story buildings without a basement, based on the damage function specific
to those building types published in USACE (2003). We based the weighting of this function on
observed empirical estimates of building heights in the Southern United States (FEMA 2011),
where 72% of residences in Texas are single-story, and 28% are 2 stories or higher. Similarly,
we made the content damage function into a composite damage function based on the same
process as structural damages. For commercial and industrial buildings, we used the depth-
damage functions published by FEMA (2011). Finally, we estimated first floor elevation values
for the various buildings in the region by creating composite substructure height for the different
25
Flood Insurance Rate Map (FIRM) zones in West Bay. We identified the FIRM zone of each
building in the region by overlaying FEMA’s preliminary Digital FIRM database information
(FEMA 2012; Figure 9) over the parcel data in the region (GCAD 2014). Then, we computed the
first floor elevation for buildings in each zone using information about the distribution of
different types of substructures in the area as well as pre/post FIRM heights from (FEMA 2011;
Chap. 3). We obtained a height of 1.52 m for A zone structures, 1.76 m for V zone structures,
and 1.17 m for all other structures.
In addition to damages, we also compute the size of the affected population by a storm by
multiplying the average population per residence by the number of damaged residential
properties. We estimated an average population of 2.34 people per residence in West Bay from
census data (USCB 2013).
Figure 8: Damage ratio curves use in the analysis.
26
Figure 9: Preliminary FEMA FIRM database (left) and parcel data value (right) used for the valuation of marshes in West Bay.
The approach adopted herein is likely to underestimate total property damages since it
does not capture losses to market driven commodities (e.g., agriculture or land on which
infrastructure is built), and property values may not have been recently calibrated to the local
market (Cannon et al. 1995). Finally, because the approach does not address people’s willingness
to pay for coastal storm protection, the avoided damages estimates obtained should not be
interpreted as the economic value of the wetlands. Instead, they represent one component of the
value of the wetlands for storm protection.
27
3.2.5 Structures
Although wetland habitats have the capability to reduce total storm water levels, it is
sometimes preferable to build levees to prevent inundation in a region of interest. However,
wetlands can still help reduce the size, and thus the construction cost, of such structures (The
Bay Institute 2013). We measure the effectiveness of wetlands at reducing the size of levees by
computing the height of a hypothetical levee, in the presence and absence of wetlands, located at
the inland limit of the marsh footprint as computed by the SLAMM model under the 1m SLR
scenario; such a levee would allow for marshes to migrate to the toe of the structure. Where this
inland limit intersected roads or developed areas, we placed this hypothetical levee on their
seaward edge (Figure 6).
For each transect, we compute the height of the hypothetical levee with 3H:1V side
slopes following the method presented in the Overtop Manual (Pullen et al. 2007, chap.6), which
relates levee height to levee slope, wave height and surge depth at the levee’s toe and the
overtopping rate 𝑞 over the structure. Assuming that the levee would be stable and only suffer
minimum damage during a particular storm, we fixed the overtopping rate at 𝑞 = 1 l/s per m
during a storm (USACE 2002, chap.VI–5).
3.2.6 Forcing Conditions
To model the protection services of wetlands, we generated a series of potential hurricane
forcing conditions. Based on an analysis of storm surge return period values in and around
Galveston Bay (Figure 3), we created a range of realistic values of open coast surge elevations
and corresponding sustained wind speed values (Table 4), using the Saffir-Simpson scale
(NOAA 2013). We defined the deep water wave characteristics created by each of these
28
synthetic hurricanes based on the range of wave height and period that those storms could
potentially generate (USACE 2002).
Table 4: Surge and offshore wave characteristics used to quantify the protective services of wetlands in Galveston Bay
Storm Name C1 C2 C3 C4 C5 Surge at Coast [m] 2.0 2.5 3.0 3.5 4.0 Sustained Wind Speed [m/s] 35.0 40.0 45.3 50.3 55.6 Offshore Wave Height [m] 7 8 8 10 11 Offshore Wave Peak Period [s] 10 11 11 12 13
We compute the total water level during each storm by linearly adding storm surge, wave
height and wave setup along each of the transects. If different transects cross paths, the total
water level at the crossing points is equal to the maximum water level of the two. Total water
level over the entire surface is then computed by linearly interpolating and smoothing maximum
total water level over the area.
As mentioned earlier, we compute the total water level generated by hurricanes by
computing the evolution of surge and waves over a series of transects drawn perpendicular to the
Bay’s shoreline, with an offshore boundary located a few meters inside the shallow Bay. To
obtain values of storm surge at that boundary, we assume that the storm surge in the Bay is equal
to the surge at the coast (Table 4). We also assign wave characteristics at that boundary using
outputs of the wave propagation model, Simulating Waves Nearshore (SWAN) (Booij et al.
1999).
We modeled the evolution of deep water hurricane waves inside the Bay using the
SWAN model on a 50 m resolution mesh, with an offshore boundary located approximately 5
km offshore and lateral boundaries located approximately 10 km away from the Bay’s
29
boundaries. For each storm in Table 4, we first deepen the whole computational domain
uniformly by the input storm surge value. Then we compute the wave characteristics in the Bay
using the offshore wave characteristics associated with each storm, assuming that the associated
wind speed blows constantly from any of the 16 cardinal and half cardinal compass directions
(Stockdon et al. 2012). Thus, for each storm, we run the SWAN model 16 times. We compute
wave forcing conditions along the Bay’s shoreline by taking the maximum wave height and
associated wave period generated by the 16 runs. Example outputs of the SWAN model for the
storm C4 with southerly winds is presented in Figure 10.
Figure 10: Example SWAN outputs for Storm C4, with constant winds blowing from south to north.
For a particular storm, we quantify the protection services delivered by wetlands by
computing the total water level over inundated areas when wetlands are present and when they
are replaced by bare land. Under current conditions, we compute total water level for each of the
forcing conditions presented in Table 4, assuming that the habitat is healthy or degraded (see
Section 3.2.3). For storm C4, we also compute total water level assuming average health
conditions using the physical parameters presented in
30
Table 3. This is then used to compute total water level for each time step in the 1 m SLR
scenario considered herein.
3.3 Blue Carbon
3.3.1 Carbon Storage and Sequestration
Wetlands are some of the most productive carbon storage and sequestering habitat types
in the world. Carbon in wetland vegetation can accumulate in four “pools”: (1) the above ground
biomass, (2) the below ground biomass, (3) the soil, and (4) the surface detrital layer. Carbon
accumulates when wetlands persist over time and ‘sequester’ it, or when non-wetlands land
cover classes convert to wetlands. On the other hand, carbon decays and is lost, or re-emitted,
when a wetland coverts to a non- wetland land cover class such as estuarine water, agricultural
land or a different type of vegetation. Hence, the amount of carbon present in wetlands at a given
time is a function of the amount and type of habitats currently present, as well as the type of
habitats that were present in the past.
We estimate changes in the amount of carbon in marshes, or blue carbon, in the first three
pools (changes in carbon content of the surface detrital layer are not captured) as sea level rises
using the InVEST Terrestrial and Blue Carbon models (Sharp et al. 2014), along with land cover
class information from SLAMM outputs. We provide a brief description of those models and
their inputs in Appendix D. Elemental carbon values in all carbon pools and accumulation and
decay rates were assigned based on values used in previous blue carbon modeling efforts in
Freeport, Texas (Walsh et al. 2013), and values found in the literature (e.g., Chmura et al. (2003),
Loomis & Craft (2010), Choi et al. (2001), or Yu et al. (2012), see also Appendix D).
Since the blue carbon model is currently designed for salt marshes, we only model carbon
accumulation and decay over time for the following three marine wetland categories: regularly
31
flooded, irregularly flooded and transitional salt marshes. Accumulation and decay that also
occur in the fresh marsh and swamp land cover classes are not incorporated into the analysis,
since the model only accounts for changes in salt marshes coverage. Furthermore, we assume
that accumulation and decay only occur in the soil pool, where the vast majority of carbon is
stored and sequestered in wetland habitats. This approach is reasonable because the amount of
carbon stored in the plant biomass remains relatively constant over time and does not accumulate
or get lost like carbon in the soil pool.
In order to take into account the wide variability in model inputs, we evaluate the relative
importance of those inputs on the amount of carbon stored and sequestered by wetlands via an
uncertainty analysis called tornado analysis (Howard 1988; Celona & McNamee 2001). To
conduct the uncertainty analysis, we first define a range of possible values (minimum, typical
value, maximum) for soil carbon content, along with carbon accumulation and decay rates. We
obtained those ranges by taking the maximum and minimum values reported in the literature that
are applicable to Gulf wetlands (see Appendix D). Next, we run the terrestrial and blue carbon
models for each of the three possible values of a single variable (e.g., soil carbon content), while
keeping all the other variables (e.g., accumulation and decay rates) constant at their typical
values. This step produces three model outputs for each variable. The variable for which outputs
vary the most is the one that drives most of the model uncertainty.
3.3.2 Valuation
In addition to computing the amount of carbon stored and sequestered by wetlands, we
also compute the economic value of sequestration (not storage) as a function of the amount of
carbon sequestered, the monetary value of each unit of carbon, and the change in the value of
carbon sequestration over time. Because local carbon emissions affect the atmosphere at a global
32
scale, we compute the economic value of sequestration using the Social Cost of Carbon (SCC)
following USIWGSCC (2010). This approach estimates the additional costs induced by climate
change caused by increased emissions of CO2, via impact on agricultural productivity, human
health, and storm frequency (U.S. EPA 2013). Since the damages incurred from carbon
emissions occur beyond the date of their initial release into the atmosphere, the damages from
emissions in any one period are the sum of future damages, discounted back to that point. Thus,
to calculate the SCC for emissions in 2050 or 2100, we computed the present value of the sum of
future damages (until 2050 or 2100 ), using a 3% discount rate (and appropriate SCC schedule in
USIWGSCC (2010) guidance).
3.4 Fisheries
Patterns of use of marshes change among fish species seasonally and with age classes
(Stunz et al. 2010). Fisheries in Galveston Bay strongly prefer saltmarshes, and especially marsh
edges (~1m shoreward from open water; (Minello & Rozas 2002; Stunz et al. 2010), rather than
shallow open water habitats, in their juvenile stage (Minello et al. 2008; Thomas et al. 1990).
We estimate the impacts of sea-level rise and marsh migration on fisheries by using two
different types of models. First we estimate changes in available marsh edge habitats for key
fisheries using the InVEST Fisheries Habitat model. Next, we estimate changes in landing of
shrimp for different sea-level rise scenarios using the InVEST Shrimp model. Note that neither
of these models has been released as part of the InVEST package.
3.4.1 Fishery Habitat Change
We assess changes in the amount of marsh edge habitats of various fisheries as a function
of changes in the location and quantity of land and marine habitats for four different
33
commercially and recreationally important fish species in Galveston Bay: (1) Blue Crab
(Callinectes sapidus), (2) Brown Shrimp (Farfantepenaeus aztecus), (3) Southern Flounder
(Paralichthys lethostigma), and (4) Red Drum (Sciaenops ocellatus). Since these species are
present in the Bay’s wetlands primarily during their larval and juvenile stages, we only consider
these life stages in our analysis (Appendix E).
Near wetlands areas, the most important and common habitat and environmental
parameters that define where these larval and juvenile fish can live are sediment type, water
depth, temperature, salinity and turbidity. Accordingly, we collected information about the
environmental and habitat preferences for each species of interest (Appendix E). However,
because of the low spatial resolution of publicly available data, and because we focus on marsh
edge regions, we only consider water depth in this analysis.
From bathymetric and topographic data, and sediment and living habitat location
information, the model first extracts the maximum livable area for a given species’ life-stage
based on depth preferences. For marsh edge areas, those are waters between -2 meters and Mean
High Water (MHW) for all species. Then the model incorporates information about location and
quantity of available living habitats for those species by generating suitable habitat maps based
on the species preferences (Appendix E). We obtained changes in the areal extent of estuarine
and wetlands areas from SLAMM outputs (WPC 2011). Lastly, using information from the
different bathymetric and environmental layers, the model computes the amount of suitable
habitat for that species. We conducted this process for the different time horizons considered,
based on projected land cover maps. Results from the model provide a first estimate of the
changes in habitat for each key species under different sea-level rise scenarios.
34
3.4.2 Shrimp Fishery Model
Shrimps life cycle is modeled using a stage-based single species population model
(Figure 11) using a Leslie matrix approach (Jensen 1974). Starting with recruitment, the model
captures the survival rate of egg/larvae to juvenile, then from juvenile to adult. These survival
rates are a function of salinity and temperature. The survival rate from juvenile to adult is also a
function of the area of marsh habitat present. The output of the model is the weight of shrimp
harvested commercially or recreationally. Harvest and natural mortality dictate recruitment. The
total area of marsh available in the Bay also influences the survival rates of shrimps at all stages.
Figure 11: Shrimp growth model.
To evaluate changes in habitat we calculated the percent change in salt marsh habitat using
the SLAMM maps for each sea-level rise scenario. We only considered changes in the SLAMM
category ‘regularly flooded marsh,’ as this is the low elevation saltmarsh used primarily by
juvenile shrimp.
35
36
4 Quantification of Ecosystem Services in the Bay
Coastal wetlands are valued by their capacity to deliver three different services: 1) coastal
protection, 2) blue carbon storage and sequestration, and 3) habitat for fisheries and food from
shrimp fishing activities. The delivery of these three services is quantified under current and
future conditions, as sea-level rises and marshes are (not) allowed to migrate landward.
4.1 Coastal Protection
In this section, the coastal protection services delivered by wetlands are quantified in
three different ways. First, we explore the value of wetlands to serve as a physical buffer
between people and storm waters. Next, we investigate the dollar value of wetlands due to their
ability to reduce water levels during storms and thereby avoid damages to structures and reduce
the number of people displaced. Finally, we discuss how the wetlands can reduce the height of
levees, if such structures were built to prevent any damages to existing structures.
4.1.1 Wetlands as Storm Buffer and Storm Water Retention Capacity
Existing shoreline structures protect people living at the water’s edge. For people living
landward of wetlands in Galveston Bay, one primary coastal protection value of this habitat is its
ability to serve as a physical buffer between them and storm waters (Costanza et al. 2008). First,
wetlands around Galveston Bay (East and West Bay, Trinity Bay) extend, on average, 6 km
inland from the shore, but can be as wide as 16 km. Additionally, with an average elevation of
approximately 1.3 m at its landward edge, wetlands protect people and property against the
impacts of tropical or subtropical storms (NOAA 2014a), which are the majority of storm types
that move through the region. Those storms typically generate surges lower than 1 m (USACE
2002), which is lower than the 7 year return surge (surge that has a 14% chance of being
exceeded in any one year) height of 1.1 m (Figure 3).
37
In addition to serving as a buffer, wetlands also deliver valuable coastal protection
services by serving as natural storm water retention systems. Currently, wetlands around
Galveston Bay can hold more than 300 million m3 of water during a storm that generates a 1 m
surge (Figure 12). As sea level rises and wetlands’ footprint migrates landward, their retention
volume also increases. However, if they are not allowed to migrate, their retention volume
decreases (Figure 12). By 2100, for example, almost 200 million m3 of water would have to be
treated or would runoff if marshes are not allowed to migrate, because of the ensuing decrease in
wetlands area.
Figure 12: Storm water retention volume by wetlands in Galveston Bay for a 1 m SLR scenario.
4.1.2 Storm Impact Reduction and Avoided Damages due to Wetlands
Another important coastal protection service delivered by wetlands is their capacity to
limit the amount of damages to properties and the number of people displaced by large storms.
This service is quantified by the wetlands’ ability to reduce storm-induced total water level,
which is obtained here by modeling storm surge elevation, wave height and wave-induced mean
water level along a series of transects drawn perpendicularly to the shoreline, along the Bay
38
(FEMA & FIS 1988; FEMA 2007a). In this section, we first discuss the ability of wetlands to
reduce total water level generated by storm C4 (Table 4) in and around the Bay, under current
and future sea-level conditions. Next, we compute the dollar value of avoided damages and the
number of people protected by wetlands for all storms considered herein.
4.1.2.1 Current Sea-Level Conditions
Under current sea-level conditions, the coastal protection value of wetlands is computed
from total water level values obtained in the presence and absence of healthy and degraded
marsh plants (Section 3.1). Water level input conditions along the Bay’s shoreline were obtained
from fixed storm surge values corresponding to the storms presented in Table 4 and from SWAN
outputs (Figure 10).
Comparison of model outputs obtained in the presence and absence of wetlands for Storm
C4 shows that marshes have the ability to reduce total water levels during storms in Galveston
Bay (Figure 13 and Figure 14). However, as indicated in the discussion of storm surge
attenuation (Section 3.2.2 and Appendix C), the protection services delivered by wetlands is a
function of the amount of habitat present, the “health” (or physical characteristics) of the plants,
and topography. Indeed, we find that healthy marshes can reduce total water level by more than
0.5 m (50%), whereas degraded habitats only reduce wave height by a few centimeters.
Additionally, even though healthy habitats are supposedly more efficient at reducing wave
energy than degraded habitats, their relative effectiveness is highly spatially variable (Figure 13).
Wetlands seem to attenuate more wave energy in the Trinity Bay region, for example, than in
West or East Bay. Also, within the West Bay region, wetlands near Chocolate Bay moderate
total water levels more than wetlands near Texas City. (Note that the isolated wave height smears
in the different maps are the results of the interpolation between different 1-D transects with
39
different lengths, which is a function of local topography.) These differences are due to small
differences in input wave height conditions at the shoreline (Figure 10), the size of the marsh
habitat in those regions, but more importantly to local differences in topography.
Figure 13: Investigation of role of wetlands at reducing total water levels during storms. Top:
Total water level during storm C4, in the absence of wetlands. Bottom Left: Difference in total
water level during storm C4 between “No Wetland” and “Healthy Wetland”. Bottom Right:
Difference in total water level during storm C4 between “No Wetland” and “Degraded Wetland”.
40
Figure 14: Example Nearshore Wave model output for a transect in Galveston Bay, in the absence and presence of vegetation.
We repeated this exercise for the storms described in Table 4. Although the difference
between outputs of the different habitat conditions varied slightly, the overall patterns and
conclusions are similar to those presented above. The only difference between the outputs for
different storms is that total water level increases with the strength of the forcing.
4.1.2.2 Sea Level Rise Scenario
To quantify the value of wetlands in Galveston Bay under a sea-level rise scenario, we
computed total water level in the Bay generated by Storm C4. It appears that as sea level rises,
storm waters reach further inland, although the overall pattern of total water distribution is
similar from one time period to another (Figure 15). Comparison of total water level values
obtained in the presence and absence of wetlands (Figure 16) shows high heterogeneity in the
potential of wetlands to reduce the impact of storm, as was observed in the previous section. Not
41
allowing wetlands to migrate also results in higher water levels, as exemplified by results in
Figure 16.
Figure 15: Total water level surface in Galveston Bay currently (2004, top), in 2050 (bottom left) and 2100 (bottom right), using forcing conditions corresponding to storm C4.
42
Figure 16: Estimates of the wetlands role in total water level attenuation, for the year 2100. Top: Total water level. Bottom Left: Difference in total water level computed in the absence and presence of wetlands, assuming that wetlands can migrate. Bottom Right: Difference between total water levels computed in the absence and presence of wetlands, assuming that wetlands do not migrate.
4.1.2.3 Valuation
For current and future wetlands coverage conditions, the protective service of wetlands are
given in dollars of avoided damages and change in the number of affected people. Results
indicate that, for current conditions, the total amount of damages and number of people affected
during storms increases with the storm intensity (Figure 17). As expected from the biophysical
results presented above, wetlands reduce the amount of damages and protect communities from
43
storm impacts. On average, healthy wetlands reduce damages by $15 million and protect 450
people whereas degraded wetlands reduce damages by $3 million and protect 105 people. For
storm C4 in particular, healthy wetlands reduce damages by $17 million and protect 348 people.
Degraded wetlands reduce damages by $4 million and protect 63 people. If we use average
physical parameters values under this storm scenario (
44
Table 3), the wetlands reduce damages by $8 million and protect 167 people, for storm C4.
Note that, although the amount of avoided damages increases with storm intensity, the number of
people protected by wetlands does not necessarily increase with storm intensity since we
assumed a fixed number of people per household, regardless of property value.
Figure 17: Top: Total amount of damages (left) and total number of people affected (right) by storms in Galveston Bay for different storms under current conditions and for different habitat conditions. Bottom: Amount of avoided damages (left) and people protected (right) by wetlands.
As sea level rises, the total amount of damages increases since storm waters extend further
inland (Figure 18). However, wetlands reduce the total amount of damages by an average of $12
million, and protect an average of 395 people if they are allowed to migrate. If marshes are not
allowed to migrate, the amount of avoided damages decreases to $9 million and the number of
people protected reduces to 310 people. Interestingly, wetlands are able to prevent more damages
as sea level rises, if they are allowed to migrate, since there is slightly more wetland area in 2050
45
than in 2100. However, if wetlands are not allowed to migrate, their total coverage declines from
2050 to 2100, yielding a lower amount of avoided damages for storms occurring in 2100
compared to 2050.
Figure 18: Same as Figure 17, for a 1 m SLR scenario.
4.1.3 Structures
Wetlands can reduce total water levels during storms, but they do not prevent damages
from occurring, as shown in the above section. Man-made protective structures can protect
people and properties against the impact of storms. But structures are also costly, especially if
they are designed to prevent damages during the strongest storms (USACE 2002).
We investigate the potential for wetlands to reduce the height of levees in West Bay by
assuming that a levee is built at the landward edge of the 2100 marsh footprint. Such a levee is
46
designed to reduce overtopping and inland inundation to a minimum. We find that, in the
absence of marsh, levee height requirements are quite spatially variable (Figure 19), as wave
height and total water levels varied in this region of the Bay (see Section 4.1.2.2). The presence
of wetlands reduces the height requirements of levees by as much as 4 to 5 m, but, again, the
potential for wetlands to reduce the height of the structure varies along West Bay. We conducted
this analysis for all the scenarios and found similar trends. In addition, we also find that levee
heights increase as sea level rises, and that wetlands are more efficient at reducing levee height if
they are allowed to migrate inland (Figure 20).
Figure 19: Effect of wetlands on levee height in West Bay, for storm C4. Left: Levee height required to allow 1 l/s per m of overtopping, in the absence of vegetation. Right: Levee height decrease due to the presence of vegetation.
47
Figure 20: Effect of wetlands on levee height in West Bay, for all storms and SLR scenarios. : Boxplot of levee heights in the presence and absence of vegetation for (a) current conditions, (b) in 2050 and (c) 2100 assuming that wetlands are (not) allowed to migrate.
4.2 Blue Carbon
We computed the total amount of elemental carbon (C) stored by salt marshes – regularly
flooded marsh, irregularly flooded marsh and transitional salt marsh – around Galveston Bay for
2004, 2050 and 2100 (Figure 21) by accounting for all the carbon stored in three different pools
1) the above ground biomass, 2) below ground biomass and 3) soil. In 2004, salt marshes stored
29.21 million metric Tons C/ha – 1 metric Ton, or Ton, is equal to 1,000 kg – of which 38% (11
million Tons C/ha) was stored within the three salt marsh habitats that make up 8% of the habitat
area. By 2050 the total amount of carbon stored increases by 1.9% to 29.76 million Tons of
carbon as marshes migrate landward. Forty-four percent (44%) of this carbon is contained in salt
marsh. However, as the rate of SLR increases in the second half of the century the amount of
carbon stored decreases by 6.2% from the initial values, a loss of over 1.8 million Tons of carbon
storage (Figure 21).
The decrease in the wetlands storage capacity during the second half of the century is a
result of salt marsh loss, and resultant carbon emissions, which exceeds the rate of creation of
48
new marsh and its ability to sequester carbon. Emission occurs as salt marsh area is inundated
and converted into open water, where the model assumes that carbon currently stored in
sediment is being re-suspended and lost to other systems. This is happening despite the fact that
salt marshes still constitute 8% of the total wetland area and the total amount of carbon stored
within salt marshes increases by almost 14% (1.5 million Tons C) between 2004 and 2100
(Figure 22). Interestingly, the amount of carbon stored in regularly flooded marshes more than
doubles during this period from 3 million Tons/C to 6.5 Million Tons/C, which offsets a loss of
carbon from irregularly flooded marsh, fresh marsh and swamp habitats of more than 3 million
Tons/C.
49
Figure 21: Carbon stored by wetlands in the Bay in 2004 (top), 2050 (bottom left) and 2100
(bottom right).
50
Figure 22: Total carbon stored by different landcover types around Galveston Bay.
We converted the changes in total carbon storage shown above to CO2 sequestration and
emission values for 2050 and 2100 (Figure 23 and Figure 24). We estimate that, between 2050
and 2100, wetlands can store 8.36 million Tons of CO2, while they can emit 6.35 million Tons.
As the rate of SLR increases between 2050 and 2100, an additional 9 million Tons CO2 are
stored, however the loss of existing marsh habitat to inundation causes emissions of CO2 to
increase to 17.6 million Tons. This results in a total predicted loss of 6.6 million Tons of CO2
from Galveston Bay habitats between 2004 and 2100.
51
Figure 23: Carbon dioxide stored and sequestered by wetlands in Galveston Bay between 2004 and 2050 (left) and 2050 and 2100 (right).
Figure 24: Carbon dioxide stored and sequestered by wetlands in Galveston Bay between 2004 2100.
The greatest losses of carbon (>99% of total emission) occur as salt marsh becomes
inundated and converts to tidal flat or estuarine water habitat. As the rate of sea level rise
52
increases between 2050 and 2100, so does the amount of carbon lost to emissions. On the other
hand, salt marshes, including those newly created by migration, account for over 93% of the CO2
sequestered in Galveston Bay habitats. It is important to remember that, in our model, the loss of
carbon from sediments occurs at a much faster rate than it can accumulate in new marsh habitat
so there is not a one-for-one exchange based on area. Indeed, we assume that salt marsh
accumulates carbon in the soil at a rate of 28.6 Tons C/ha over a 50 year period, while the loss of
regularly flooded marsh emits 189 Tons C/ha in the same time period. In summary, we find that,
after an initial gain of carbon between 2004 and 2050, the Bay becomes a net emitter of carbon
between 2050 and 2100, for a net emission of CO2 of -6.6 million Tons through the year 2100
(Figure 25). Interestingly, DeLaune & White (2011) found similar results in Louisiana.
Figure 25: Amount of CO2 sequestered (positive values) and emitted (negative values) by wetlands for different time horizons. Total amount of carbon stored by wetlands is equal to the sum of carbon sequestered and emitted by wetlands.
We performed the same analysis assuming that marshes were not allowed to migrate.
Under this assumption, all new marsh land that was created by SLAMM is removed, but existing
salt marshes are allowed to convert to other types of salt marsh habitat within the initial marsh
footprint (i.e., from irregularly flooded to regularly flooded). Also, carbon only accumulates
where marsh persists, and is lost otherwise. Results indicate that in 2004 the total amount of
53
carbon stored was approximately 29.21 million Tons, 38% of which was stored within the three
salt marsh habitats. By 2050, the total amount of carbon in the project area decreases by
approximately 0.9% to 28.96 million Tons of carbon. This decrease is due to the fact that salt
marsh area also decreases, as wetlands do not migrate landward. By 2100, where the rate of SLR
increases, the amount of carbon stored is projected to be reduced by 7.7% from the initial values,
a loss of over 2.2 million Tons C. Note that, since the fresh marsh and swamps habitats are not
affected by the “No Migration” restriction, the total reduction in storage capacity is caused by
changes in salt marsh habitats, tidal flats, and estuarine waters areas. Without marsh migration
salt marsh habitats lose over 10% (1.1 million Tons C) of their initial carbon by 2050, while
under the marsh migration scenario they gain 20% (2.2 million Tons C). By 2100, regularly
flooded marsh habitat is predicted to gain only 0.5 million Tons C while both irregularly flooded
marsh and transitional salt marsh are predicted to lose 3.4 million Tons C and 1.1 million Tons
C, respectively. In summary, under the “no marsh migration” scenario, the Bay emits 0.9 million
Tons CO2 between 2004 and 2050. Between 2050 and 2100, an additional 7.4 million Tons of
CO2 are emitted, for a total net loss of 8.3 million Tons of CO2 (Figure 25).
We translate those estimates of carbon storage and sequestration into net present value of
avoided (positive value) or incurred (negative value) climate damages from carbon
sequestration/emissions as a result of landcover change from 2004-2100, given in 2010 US
dollars. If wetlands are allowed to migrate, they avoid approximately $3 M in climate damage by
sequestering carbon between 2004 and 2050 (Figure 26). However, as sea level continues to rise,
the bay incurs climate damages of up to $11 M by 2100. Those environmental damages are even
greater if marshes are not allowed to migrate, with $23 M in emissions-related damages by 2050
and $31 M of damages by 2100.
54
Figure 26: Climate damage from carbon sequestration and emission by marshes in Galveston Bay.
These results indicate that, unless sea level rises slowly, the Bay might become a net
emitter of carbon before the end of the century, resulting in significant damages. To gain more
insight into the way in which the rate of sea level rise might impact the level of damages incurred
by wetlands, we computed the amount of carbon stored and sequestered by the Bay assuming a
0.4 m sea level rise (IPCC mean) and 2 m rise by 2100, which we converted into the amount of
damages avoided or incurred (Figure 27). We find that the amount of damages incurred increases
dramatically to $86 M if sea level rises by 2 m in 2100. However, if sea level rises moderately to
0.32 m by 2100 (which is lower than current estimates of 0.7 m), wetlands avoid more than $20
M of damages.
55
Figure 27: Climate damage from carbon sequestration and emission by marshes in Galveston Bay for different sea level rise scenarios.
As argued in Section 3.2, inputs to the blue carbon model vary widely from region to
region. The inputs that we chose are the results of an educated guess that might not reflect the
exact (and still unknown) carbon storage and decay rates of habitats in the Bay. In order to
determine the relative importance of (1) soil organic carbon content in salt marsh, (2) carbon
accumulation rates, and (3) decay rates, we conducted an uncertainty analysis, using a tornado
approach, for the 1 m SLR scenario. Table 5 shows the range of values used for each of the input
parameters for the tornado analysis.
Table 5: Range of values for the different blue carbon model input parameters used in the uncertainty analysis.
Model Parameter Minimum Model Input Maximum Soil carbon content [Tons C/ha] 100 222 462 Accumulation [Tons C/ha] 0.18 0.57 8.22 Decay Rate [%] Tidal Flat 0.25 0.50 0.75 All other land cover types 0.50 0.75 1.0
We find that when soil carbon content or the decay rates are set at their minimum values,
the Bay emits approximately 10 thousand more Tons of carbon compared to when we were using
56
the default input values. When those two parameters are set at their maximum values, the Bay
stores a few thousand more Tons of carbon. However, if the accumulation rate is set at a
minimum, while the other two remain at their default input value, the Bay stores more than 150
thousand tons of carbon. If this parameter is set at its maximum value, the bay becomes a net
emitter of carbon.
This analysis shows that any estimate of the carbon storage and sequestration potential of
wetlands will contain high uncertainty, as mentioned by Lenart (2009). Also, although the exact
value of the different input parameters is important to correctly estimate the storage potential of
wetlands, the accumulation rate value is the most important input. Changes in its value yield the
largest swings in carbon storage capacity.
Figure 28: Results of the blue carbon model uncertainty analysis. Results are computed for the period 2004 to 2100.
4.3 Fisheries
In this section, we quantify the impacts of SLR and changes in marsh habitat coverage on
the selected fisheries in Galveston Bay in two distinct ways. First, we investigate how an
57
increase in sea level changes the amount of key habitats used by the juvenile fish species
described in Section 3.4.1. Next, we quantify how changes in marsh habitats coverage due to
SLR can translate into changes in landings of shrimp.
4.3.1 Fisheries Habitat Change
Under current conditions, there are nearly 280 km2 of suitable nearshore habitats
(wetlands, tidal flats and nearshore estuarine waters) for larval and juvenile blue crabs, brown
shrimp, southern flounder and red drum in Galveston Bay (Appendix F). The total footprint of
those habitats will increase as sea-level rises, especially tidal and estuarine water areas.
Additionally, under current conditions, larval and juvenile blue crabs, brown shrimp, southern
flounder and red drum can utilize nearly 1.4 km2 of marsh edge area in Galveston Bay (Figure 29
and Figure 30). As sea level rises and if wetlands are allowed to migrate, the location of marsh
edge used shifts landward, and the marsh edge area also increases to nearly 2.4 km2 (a 78%
increase) through 2050 and 4.5 km2 (a 240% increase) by 2100. This increase is mostly due to
favorable changes in topography, where future edge areas are on flatter surfaces and thus have a
greater surface area than the current edge region.
58
Figure 29: Marsh edges for the years 2004, 2050 and 2100
Figure 30: Detail of marsh edge spatial extent in West Bay for (A) all three time steps, (B) between 2004 and 2050 and (C) between 2050 and 2100.
4.3.2 Shrimp Population Growth
We investigate how changes in marsh footprint with sea-level rise translate into landing
increases of brown shrimp using the shrimp population model described in Section 3.4.2. Model
A
B C
59
results indicate that, as anticipated from the habitat modeling results presented in the previous
section, the amount of shrimp caught increases as sea level rises, since more marsh habitat is
created (Figure 31). Specifically, shrimp catch might increase by 39% by 2050, and by more than
100% by 2100. However, shrimp catch may decrease if marsh area is lost. For example, a 10%
loss in wetlands translates into a 7% loss in shrimp catch, and a 20% loss of marsh habitat results
in a 17% reduction in the shrimp commercial catches.
Figure 31: Changes in Shrimp landing as a function of changes in marsh habitats.
5 Protecting Galveston Bay Wetlands
In this document, we have shown that wetlands have the potential to reduce the impacts
of storms by reducing damages to properties and by protecting people. Wetlands also store and
sequester carbon and provide habitats for fisheries. However, we have also shown that a rise in
sea level will create more damages to structures and might cause more damage to the
60
environment by increasing CO2 emissions. In this section, we examine how oyster reefs can
minimize the impacts of SLR on the Bay’s coastline and in the process minimize the loss of
wetlands.
Oyster reefs have been shown to moderate nearshore wave energy and protect marshes
against wave-induced erosion in the Gulf (Kroeger & Guannel 2014; Scyphers et al. 2011).
Similar to artificial submerged breakwaters, oyster reefs have the ability to break nearshore
waves and generate a relatively quiet wave climate in nearshore regions. Figure 32 (below)
shows a profile of wave height over a 1-D transect in Galveston Bay, in the presence and absence
of an oyster reef. The reef reduces the wave height in the nearshore by more than 10 cm. Thus,
even though marshes can still be affected by sea-level rise, conveniently situated oyster reefs
might help reduce shoreline erosion and the loss of marshes by reducing the number of stressors
on wetlands (WPC 2009). This beneficial effect of oyster reefs can last for decades, since oyster
reefs can grow vertically as sea-level rises (Rodriguez et al. 2014) and thus maintain the low
wave energy climate nearshore.
Roland & Douglass (2005) showed that marsh erosion in Mobile Bay, AL, is partially
dictated by the relative height of wind-generated waves in the Bay. Larger waves generate more
erosion than smaller waves. To evaluate the potential for reefs to protect marshes in the Bay, we
generated a time series of waves in the Bay following the method of Roland & Douglass (2005),
using wind speed measured at Pleasure Pier and an average depth of 2.5 m over the profile.
Then, we modeled the evolution of those waves on the depth profile shown in Figure 32, in the
presence and absence of a reef. In general, reefs can reduce the height of the highest waves in the
Bay, and thus reduce their impacts on the wetlands (Figure 33).
61
Figure 32: Profile of wave height reduction in the presence of oyster reef on an arbitrary profile in Galveston Bay.
Figure 33: Exceedence profile of wind-wave height in proximity of marshes in the presence and absence of an oyster reef.
62
Based on this result, we developed an oyster restoration suitability map (Figure 34). This
map shows the 80th percentile wave height in different regions of the Bay. According to Roland
& Douglass (2005), marshes exposed to H80 greater than 0.2 m are more susceptible to erosion
than others. Consequently, Figure 34 shows that oyster reef might provide more services in East
and Trinity Bay regions than in West Bay.
Figure 34: Values of 80th percentile wind-generated wave height inside Galveston Bay.
63
6 Conclusion
In this document, we quantified the coastal protection, carbon storage and sequestration
services, and fisheries services delivered by wetlands in Galveston Bay. We performed this
analysis using a marsh footprint representing conditions in 2004 and future marsh footprints for
2050 and 2100, as predicted by the model SLAMM. Those future wetland maps were generated
assuming that sea level rises by 1 m through 2100, compared to the 1992 baseline.
We find that wetlands have the capacity to store millions of cubic meters of storm waters,
and thus provide a buffer to communities. They also attenuate total water levels during storms in
the Bay, and they can avoid millions of dollars in property damage and protect hundreds of
people against the impact of storms. However, their ability to moderate the impacts of storms is
highly spatially variable, dependent on local topography and on the choice of physical
characteristics of the plants. Wetlands can also reduce the construction costs of structures by
reducing their height requirements to prevent any overtopping during storms. Those services can
continue to be delivered as sea-level rises, as long as wetlands are allowed to migrate and are in
relatively healthy conditions. If wetlands are not allowed to migrate, their coastal protection
value decreases as sea level rises.
Additionally, we find that currently, wetlands in the Bay avoid millions of dollars of
environmental damages by sequestering carbon. The Bay can continue to sequester carbon as sea
level rises, as long as that rate of increase is relatively slow. If sea level rises by 1 m in 2100,
which is likely given current local estimates, hundreds of millions in environmental damages
may result from the release of more CO2, by submerged carbon-rich sediments, than wetlands
can store and sequester.
64
Finally, we find that most fisheries species might benefit from larger areas of available
habitats as sea level rises. Additionally, for brown shrimp, an increase in marsh habitat translates
into net gains in landing. Thus, it is likely that all commercial and recreational fisheries activities
in the Bay might benefit from a rise in sea level, if marshes are allowed to migrate.
However, as mentioned in the main text, this analysis contains some uncertainty. For
example, we used average estimates of habitat conditions and physical characteristics around the
Bay to compute the coastal protection services delivered by wetlands. We also assumed that the
location, type and value of properties will not change through time. Better habitat data and more
accurate predictions of property owners’ behaviors are likely to improve the protection service
estimates of wetlands. Furthermore, the evaluation of carbon storage and sequestration is highly
sensitive to the plants accumulation rate, which is an important unknown characteristic in the
Bay. There is also a fair amount of uncertainty in the fate of carbon released by sediments once
submerged wetlands disappear. As mentioned earlier, despite the advances made herein, there
are still many research gaps to accurately quantify the carbon storage and sequestration capacity
of wetlands (Lenart 2009). Finally, despite the fact that we find that fisheries might benefit from
more habitats as sea-level rises, fisheries stand to be directly impacted to a much greater degree
by changes in temperature and salinity, two variables not included in our analysis. These two
environmental factors regulate key life cycle events for many marine species (Copeland 1966;
Anger 2003; Ward 2012; Hildebrand & Gunter 1953; Zein-Eldin & Renaud 1986; Buzan et al.
2009; Needham et al. 2012) and may have a greater impact on fisheries than any increase in
habitat area caused by sea level rise.
As a consequence, results can be interpreted as an estimate of the storm damage
mitigation services provided by wetlands and their capacity to improve fisheries or store carbon.
65
Nevertheless, our results and approach is one of the first attempts at a comprehensive accounting
of different services delivered by wetlands in the face of rising seas. We show that an increase in
sea level might increase the impacts of storms and degrade the environment by increasing
releases of CO2 in the atmosphere. Preserving wetlands and allowing them to migrate naturally
can reduce the impacts of those stressors and might improve fisheries in the Bay.
66
7 Future Work
This study has demonstrated that wetlands in Galveston Bay have the potential to
moderate the impacts of storms currently, and in the future, as sea-level rises. It also
demonstrated that wetlands can sequester large amount of carbon and sustain fisheries. However,
this study has also identified a series of research needs to better quantify the services delivered
by wetlands.
The quantification of ecosystem services by wetlands requires an accurate accounting of
the type and location of different species of marsh plants in and around Galveston Bay under
different sea-level rise scenarios. Herein, we used SLAMM outputs for a 1 m SLR scenario,
which contain some inaccuracies as SLAMM predicts that non-developed and developed areas
will be converted to marsh in 2050 and 2100. It is unlikely that such land conversions will occur
based on discussions with local partners. More accurate predictions of marsh footprint at
different time horizons will improve the quantification of the changes in the delivery of services
by wetlands.
The estimates of total water level in the presence of wetlands might be improved by using
a 2D model for wave and storm surge evolution, such as SWAN and/or ADCIRC (Luettich et al.
1992; Booij et al. 1999). However, this analysis has shown the need for better representations of
the physical parameters of plants in order to improve the accuracy of the model outputs. In
addition, the valuation of the protective role of wetlands might be improved with more
information about the location and characteristics of the different structures and the composition
of the households around Galveston Bay.
67
Estimates of the carbon storage and sequestration capacity of the wetlands might be
improved with the inclusion of more accurate information about the ability of the soil and marsh
plants in and around the Bay to store carbon. The model used in the analysis might also be
improved by considering non-linear rates of accumulation or loss.
Finally, the fisheries models might be improved by including the role of more
environmental attributes (i.e., sediment type distribution, salinity and temperature data) in both
the habitat and population models. In particular, we did not include any information about how
temperature or salinity will change in 2050 and 2100 compared to the current baseline.
Additionally, any future work should include estimates of population changes for other important
species in the Bay, such as blue crabs, red drums, etc.
68
8 Acknowledgements
This research was funded by the National Oceanic and Atmospheric Administration through their Grant NA11OAR4310136. We would like to thank J. Blackburn and the LSCNRA Project Team, P. Bedient, C. Dawson, R. Sass, B. Stokes President of the Galveston Bay Foundation and the staff of the Houston Wilderness.
9 References
Agnew, M., 2012. Surge and Wave Propagation over Wetlands with Respect to Storm Forward Speed. University of Notre Dame, Indiana.
Anderson, M.E., Smith, J.M. & Mckay, S.K., 2011. Wave Dissipation by Vegetation; Report ERDC/CHL CHETN I 82, Vicksburg, MS.
Anger, K., 2003. Salinity as a key parameter in the larval biology of decapod crustaceans. Invertebrate Reproduction & Development, 43(1), pp.29–45. Available at: http://www.tandfonline.com/doi/abs/10.1080/07924259.2003.9652520.
Arnold, C. R., W. H. Bailey, T. D. Williams, A. Johnson & Lasswell, J.L., 1977. Laboratory spawning and larval rearing of red drum and southern flounder. Proc. Southeast. Assoc. Fish Wildl. Agencies, 31, pp.437–440.
Augustin, L.N., Irish, J.L. & Lynett, P., 2009. Laboratory and numerical studies of wave damping by emergent and near-emergent wetland vegetation. Coastal Engineering, 56(3), pp.332–340.
Van Den Avyle, M.J. & Fowler.., D.L., 1984. Species profiles: life histories and environmental requirements of coastal fishes and invertebrates (south Atlantic): blue crab. U.S. Fish and Wildlife Service Biological Services Program, FWS/OBS 82, p.16.
Barrett, B.B. & Ralph, E.J., 1976. 1976 Environmental conditions relative to shrimp production in coastal Louisiana. La. Dept. Wildlife and Fishery Technical Bulletin, 21, p.20.
Blake, E.S., Landsea, C.W. & Gibney, E.J., 2011. The deadliest, costliest, and most intense United States tropical cyclones from 1851 to 2010 (and other frequently requested hurricane facts), Miami, FL.
Bohannon, C., 2013. 1981-2010 Commercial landings and ex-vessel value of all commercial fished species in Texas bays and offshore fishery.
Booij, N., Ris, R.C. & Holthuijsen, L.H., 1999. A third-generation wave model for coastal regions 1. Model description and validation. Journal of Geophysical Research, 104(C4), pp.7649–7666.
69
Bozek, C.M. & Burdick, D.M., 2005. Impacts of Seawalls on Saltmarsh Plant Communities in the Great Bay Estuary, New Hampshire USA. Wetlands Ecology and Management, 13(5), pp.553–568.
Buckley, J., 1984. Habitat suitability index models: larval and juvenile red drum. U.S. Fish and Wildlife Service, FWS/OBS-82, p.15.
Bunya, S. et al., 2010. A High-Resolution Coupled Riverine Flow, Tide, Wind, Wind Wave, and Storm Surge Model for Southern Louisiana and Mississippi. Part I: Model Development and Validation. Monthly Weather Review, 138(2), pp.345–377. Available at: http://journals.ametsoc.org/doi/abs/10.1175/2009MWR2906.1 [Accessed May 15, 2012].
Burkett, V. & Kusler, J., 2000. Climate change: potential impacts and interactions in wetlands of the United States. Journal of the American Water Resources Association, 36(2), pp.313–320.
Buzan, D. et al., 2009. Positive Relationship between Freshwater Inflow and Oyster Abundance in Galveston Bay, Texas. Estuaries and Coasts, 32(1), pp.206–212. Available at: http://link.springer.com/10.1007/s12237-008-9078-z.
Cannon, M.G., Phelan, J.M. & Passaro, M.A., 1995. Procedural guidelines for estimating residential and business structure value for use in flood damage estimations; IWR Report 95-R-9, Alexandria, VA.
Celona, J. & McNamee, P., 2001. Decision Analysis for the Professional 4th ed. M. Campbel, B. Roehl, & M. Story, eds., SmartOrg Inc.
Chmura, G.L. et al., 2003. Global carbon sequestration in tidal, saline wetland soils. Global Biogeochemical Cycles, 17(4), p.n/a–n/a.
Choi, Y. et al., 2001. Vegetation succession and carbon sequestration in a coastal wetland in northwest Florida: Evidence from carbon isotopes. Global Biogeochemical Cycles, 15(2), pp.311–319.
Christmas, J.Y., Langley, W. & Devender, T. Van, 1976. Investigations of commercially important penaeid shrimp in Mississippi.,
Copeland, B.J., 1966. EFFECTS OF DECREASED RIVER FLOW ON ESTUARINE ECOLOGY. Journal (Water Pollution Control Federation), 38(11), pp.1831–1839. Available at: http://www.jstor.org/stable/25035675.
Costanza, R. et al., 2008. The Value of Coastal Wetlands for Hurricane Protection. Ambio, 37(4), pp.241–248.
70
Craft, C. et al., 2009. Forecasting the effects of accelerated sea-level rise on tidal marsh ecosystem services. Frontiers in Ecology and the Environment, 7(2), pp.73–78. Available at: http://www.esajournals.org/doi/abs/10.1890/070219.
Dahl, T.E., 2011. Status and trends of wetlands in the conterminous United States 2004 to 2009, Washington, D.C .
DeLaune, R.D. & White, J.R., 2011. Will coastal wetlands continue to sequester carbon in response to an increase in global sea level?: a case study of the rapidly subsiding Mississippi river deltaic plain. Climatic Change, 110(1-2), pp.297–314.
Dugan, J.E. & Hubbard, D.M., 2009. Ecological Effects of Coastal Armoring: A Summary of Recent Results for Exposed Sandy Beaches in Southern California. In H. Shipman et al., eds. Puget Sound Shorelines and the Impacts of Armoring — Proceedings of a State of the Science Workshop. Reston, Va.: U.S. Geological Survey Scientific Investigations Report 2010–5254, p. 278.
Elsner, J.B., Jagger, T.H. & Liu, K.., 2008. Comparison of hurricane return levels using historical and geological record. Journal of Applied Meteorology and Climatology, 47, pp.368–374.
Enge, K.M. & Mulholland, R., 1985. Habitat suitability index models: southern and gulf flounders. U.S. Fish and Wildlife Service, National Coastal Ecosystems Team, Biological Rept., 82(10.92), p.38.
Erwin, K.L., 2008. Wetlands and global climate change: the role of wetland restoration in a changing world. Wetlands Ecology and Management, 17(1), pp.71–84.
Evink, G.L., 1976. Some aspects of the biology of the blue crab, Callinectes sapidus Rathbun, on Florida’s gulf coast. University of Florida, Graduate Council, Thesis, p.76 pp.
Feagin, R.A. et al., 2010. Salt Marsh Zonal Migration and Ecosystem Service Change in Response to Global Sea Level Rise : A Case Study from an Urban Region. Ecology and Society, 15(4).
Feagin, R.A. et al., 2011. Short communication: Engineering properties of wetland plants with application to wave attenuation. Coastal Engineering, 58(3), pp.251–255.
Federal Emergency Management Agency (FEMA), 2007a. Atlantic Ocean and Gulf of Mexico Coastal Guidelines Update - Final Draft, Washington, D.C. Available at: http://www.fema.gov/media-library-data/1388780453134-c5e577ea3d1da878b40e20b776804736/Atlantic+Ocean+and+Gulf+of+Mexico+Coastal+Guidelines+Update+(Feb+2007).pdf.
Federal Emergency Management Agency (FEMA), 2012. Flood Insurance Study - GALVESTON COUNTY, TEXAS, AND INCORPORATED AREAS; Vols 1-4, Washington, D.C.
71
Federal Emergency Management Agency (FEMA), 2007b. Guidelines for Coastal Flood Hazard Analysis and Mapping for the Pacific Coast of the United States, Washington, D.C. Available at: http://www.fema.gov/national-flood-insurance-program-flood-hazard-mapping/guidelines-coastal-flood-hazard-analysis-and.
Federal Emergency Management Agency (FEMA), 2011. Multi-hazard loss estimation methodology: Flood model HAZUS-MH; FEMA Technical Manual, Washington, D.C.
Federal Emergency Management Agency (FEMA) & Federal Insurance Administration (FIS), 1988. Wave Height Analysis for Flood Insurance Studies - Technical Documentation for Whafis Program Version 3.0,
Galveston Central Appraisal District (GCAD), 2014. 2014 Galveston County Shape Files. Available at: http://www.galvestoncad.org/PA/Shapeidx/shapes.htm [Accessed January 1, 2014].
Ginsburg, I., 1952. Flounders of the genus Paralichthys and related genera in American waters. Fishery Bulletin, 52(71), pp.267–351. Available at: http://fishbull.noaa.gov/52-1/ginsburg.pdf.
Glass, L. a. et al., 2008. Distribution, condition, and growth of newly settled southern flounder (Paralichthys lethostigma) in the Galveston Bay Estuary, TX. Journal of Sea Research, 59(4), pp.259–268. Available at: http://linkinghub.elsevier.com/retrieve/pii/S1385110108000105.
Guannel, G. et al., 2014. Quantification of Coastal Protection Services Delivered by Vegetation. Journal of Geophysical Research.
Gunter, G., Christmas, J.Y. & Killebrew, R., 1964. Some relations of salinity to population distributions of motile estuarine organisl~s with special reference to penaeid shrimp. Ecology, 45, pp.181–185.
Gutherz, L.J., 1967. Field guide to the flatfishes of the family Bothidae in the western north Atlantic. U.S. Fish Wildl. Servo Circ., 263, p.47.
Haby, M.G., Miget, R.J. & Falconer, L.L., 2009. Hurricane Damage Sustained by the Oyster Industry and the Oyster Reefs Across the Galveston Bay System with Recovery Recommendations, Available at: http://texas-sea-grant.tamu.edu/WhoWeAre/AgentsSpecialist/mike_haby/Economic-Damages_Galveston-Bay-Oyster-Industry_copy-ready.pdf.
Hallegatte, S. et al., 2013. Future flood losses in major coastal cities. Nature Clim. Change, 3(9), pp.802–806. Available at: http://dx.doi.org/10.1038/nclimate1979.
72
Hammerschmidt, P., 1982. Life history of the blue crab, Callinectes sapidus Rathbun, along the Texas coast. In: Proceedings of the blue crab colloquium (H. Perry and W. Van Engel, eds.), 13-14., Tech. Rep.
Van Heukelem, W.F., 1991. Blue Crab - Callinectes sapidus. In D. Funderburk, S.L., Mihursky, J.A., Jordan, S.J., Riley, ed. Habitat Requirments for Chesapeake Bay Living Resources. Chesapeake Bay Program, p. 24.
Hildebrand, H.H. & Gunter, G., 1953. Correlation of Rainfall with the Texas Catch of White Shrimp, Penaeus Setiferus (Linnaeus). Transactions of the American Fisheries Society, 82(1), pp.151–155.
Houston-Galveston Area Council (H-GAC), 2013. 2040 Regional Growth Forecast. Available at: http://www.h-gac.com/community/socioeconomic/2040-regional-growth-forecast/default.aspx [Accessed January 1, 2013].
Howard, R.A., 1988. Decision Analysis: Practice and Promise. Management Science, 34(6), pp.679–695.
IPCC, 2013. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change T. F. Stocker et al., eds., Cambridge, United Kingdom and New York, NY, USA: Cambridge University Press.
Jadhav, R.S., Chen, Q. & Smith, J.M., 2013. Spectral distribution of wave energy dissipation by salt marsh vegetation. Coastal Engineering, 77(null), pp.99–107.
Jensen, A.L., 1974. Leslie Matrix Models for Fisheries Studies. Biometrics, 30(3), pp.547–551. Available at: http://www.jstor.org/stable/2529208.
Keim, B.D., Muller, R.A. & Stone, G.W., 2007. Spatiotemporal Patterns and Return Periods of Tropical Storm and Hurricane Strikes from Texas to Maine. Journal of Climate, 20(14), pp.3498–3509.
Kroeger, T. & Guannel, G., 2014. Fishery enhancement and coastal protection services provided by two restored Gulf of Mexico oyster reefs. In K. N. Ninan, ed. Valuing Ecosystem Services-Methodological Issues and Case Studies. Northampton, MA: Edward Elgar Publishers, p. 464.
Leatherman, S.R., 1984. Coastal Geomorphic Responses to Sea Level Rise: Galveston Bay, Texas. In M. C. Barth & J. G. Titus, eds. Greenhouse Effect and Sea Level Rise: A Challenge for this Generation. Washington, D.C.: Van Nostrand Reinhold Company Inc.
Lenart, M., 2009. An unseen carbon sink. Nature Climate Change Reports, (0912), pp.137–138.
73
Lester, L.J. & Gonzalez, L.A., 2011. The State of the Bay: A Characterization of the Galveston Bay Ecosystem, 3rd Ed., Houston, TX. Available at: http://www.galvbaydata.org/stateofthebay/tabid/1846/default.aspx.
Levin, P.S. & Stunz, G.W., 2005. Habitat triage for exploited fishes: Can we identify essential “Essential Fish Habitat?” Estuarine, Coastal and Shelf Science, 64(1), pp.70–78. Available at: http://linkinghub.elsevier.com/retrieve/pii/S0272771405000405.
Loomis, M.J. & Craft, C.B., 2010. The fumigation extraction method to estimate soil microbial biomass: calibration of the KEN value. , pp.1028–1036.
Luettich, R.A., Westerink, J.J. & Scheffner, N.W., 1992. ADCIRC: an advanced three-dimensional circulation model for shelves coasts and estuaries, report 1: theory and methodology of ADCIRC-2DDI and ADCIRC-3DL, Dredging Research Program Technical Report DRP-92-6,
Mattheus, C.R. et al., 2010. Impact of land-use change and hard structures on the evolution of fringing marsh shorelines. Estuarine, Coastal and Shelf Science, 88(3), pp.365–376.
Mattocks, C. & Forbes, C., 2008. A real-time, event-triggered storm surge forecasting system for the state of North Carolina. Ocean Modelling, 25(3-4).
McEachran, J.D. & Fechhelm, J.D., 2006. Fishes of the Gulf of Mexico Volume 2., University of Texas Press.
McGee, B.D. et al., 2006. Hurricane Rita Surge Data, Southwestern Louisiana and Southeastern Texas, September to November 2005, Reston, Va. Available at: http://pubs.usgs.gov/ds/2006/220/.
Mendez, F.J. & Losada, I.J., 2004. An empirical model to estimate the propagation of breaking and nonbreaking waves over vegetation fields. Coastal Engineering, 51, pp.103–118.
Miles, J.R., Russell, P.E. & Huntley, D.A., 2001. Field Measurements of Sediment Dynamics in Front of a Seawall. Journal of Coastal Research, 17(1), pp.195–206.
Millennium Ecosystem Assessment, 2005. Ecosystems and Human Well-being: Synthesis, Washington, D.C.
Minello, T.J. et al., 2008. Population and Production Estimates for Decapod Crustaceans in Wetlands of Galveston Bay, Texas. Transactions of the American Fisheries Society, 137(1), pp.129–146. Available at: http://www.tandfonline.com/doi/abs/10.1577/T06-276.1.
Minello, T.J. & Rozas, L.P., 2002. Nekton in gulf coast wetlands: fine-scale distributions, landscape patterns, and restoration implications. Ecological Applications, 12(2), pp.441–455. Available at: http://www.esajournals.org/doi/abs/10.1890/1051-0761(2002)012[0441:NIGCWF]2.0.CO;2.
74
National Oceanic and Atmospheric Administration (NOAA), 2014a. Historical Hurricane Tracks. Available at: http://maps.csc.noaa.gov/hurricanes/# [Accessed January 1, 2014].
National Oceanic and Atmospheric Administration (NOAA), 2014b. Regional Mean Sea Level Trends. Tides & Currents. Available at: http://tidesandcurrents.noaa.gov/sltrends/slrmap.html [Accessed January 1, 2014].
National Oceanic and Atmospheric Administration (NOAA), 2013. Saffir-Simpson Hurricane Wind Scale. Available at: http://www.nhc.noaa.gov/aboutsshws.php [Accessed January 1, 2014].
Needham, H., Brown, D. & Carter, L., 2012. Impacts and adaptation options in the gulf coast, Available at: http://www.c2es.org/docUploads/gulf-coast-impacts-adaptation.pdf.
Needham, H.F. & Keim, B.D., 2011. A storm surge database for the US Gulf Coast. International Journal of Climatology, 32(14), pp.2108–2123.
NMFS, 2010. NOAA, NMFS Commercial Fisheries Statistics. Available at: http://www.st.nmfs.noaa.gov/commercial-fisheries/.
NOAA, 2012. Fisheries Economics of the United States 2012. , 2012. Available at: https://www.st.nmfs.noaa.gov/economics/publications/feus/fisheries_economics_2012#.
NOAA, 1998. Generic Amendment for Addressing Essential Fish Habitat Requirements in the following Fishery Management Plans of the Gulf of Mexico: Shrimp Fishery..... Gulf of Mexico Fishery Management Council, October 19, p.244.
NOAA Coastal Services Center (NOAA-CSC), 2006. C-CAP Land Cover Atlas. Available at: http://www.csc.noaa.gov/digitalcoast/tools/lca [Accessed January 1, 2014].
NOAA Coastal Services Center (NOAA-CSC), 2013. Datums for 8771510, Galveston Pleasure Pier TX. NOAA Tides & Currents. Available at: http://tidesandcurrents.noaa.gov/datums.html?id=8771510.
NOAA National Buoy Data Center (NDBC), 2014. Station 42035 (LLNR 1145) - GALVESTON,TX. Available at: http://www.ndbc.noaa.gov/station_page.php?station=42035 [Accessed January 1, 2014].
NOAA Office of Response and Restoration, 2012. Environmental Sensitivity Index (ESI) Maps. Available at: http://response.restoration.noaa.gov/maps-and-spatial-data/environmental-sensitivity-index-esi-maps.html [Accessed June 1, 2012].
Orth, R.J. & van Montfrans, J., 1990. Utilization of Marsh and Seagrass Habitats by Early Stages of Callinectes Sapidus: A Latitudinal Perspective. Bulletin of Marine Science, 46(1), pp.126–144. Available at:
75
http://www.ingentaconnect.com/content/umrsmas/bullmar/1990/00000046/00000001/art00011.
Packer, D.B. et al., 1999. Essential Fish Habitat Source Document: Summer Flounder, Paralichthys dentatus, Life History and Habitat Characteristics. NOAA Technical Memorandum, NMFS-NE-15, p.98.
Pardieck, R. et al., 1999. Ontogenetic changes in habitat use by postlarvae and young juveniles of the blue crab. Marine Ecology Progress Series, 186, pp.227–238. Available at: http://www.int-res.com/abstracts/meps/v186/p227-238/.
Park, R.A., Armentano, T.V. & Cloonan, C.L., 1986. Predicting the Effects of Sea Level Rise on Coastal Wetlands. In J. G. Titus, ed. Effects of Changes in Stratospheric Ozone and Global Climate, Vol. 4: Sea-Level Rise. Washington, D.C.: U.S. Environmental Protection Agency, pp. 129–162.
Perret, W.S., Weavers, J.E., Williams, R.O., Johansen, P.L., McIlwain, T.D., Raulerson, R.C. & Tatum, W.M., 1980. Fishery profiles of red drum and spotted seatrout. Gulf States Marine Fisheries Commission, 6, p.66.
Perry, H. & Stuck, K., 1982. The life history of the blue crab in Mississippi with notes on larval distribution. In: Proceedings of the blue crab colloquium (H. Perry and W. Van Engel, eds.), 17-22. T, Tech. Rep.
Perry, H.M. & McIlwain, T.D., 1986. Species Profiles: Life Histories and Environmental Requirements of Coastal Fishes and Invertebrates (Gulf of Mexico): Blue Crab. U.S. Department of the Interior, pp.1–31.
Port of Houston Authority, 2012. The Port - Overview. Available at: http://www.portofhouston.com/about-us/overview/ [Accessed January 1, 2014].
Pullen, T. et al., 2007. Eurotop: Wave Overtopping of Sea Defences and Related Structures; Assessment Manual, Available at: www.overtopping-manual.com.
Rice, H., 2013. Sea swallowing Galveston faster than thought. Houston Chronicle. Available at: http://www.houstonchronicle.com/news/houston-texas/houston/article/Sea-swallowing-Galveston-faster-than-thought-4160081.php [Accessed September 23, 2014].
Rodriguez, A.B. et al., 2014. Oyster reefs can outpace sea-level rise. Nature Climate Change, 4(6), pp.493–497.
Roland, R.M. & Douglass, S.L., 2005. Estimating Wave Tolerance of Spartina alterniflora in Coastal Alabama. Journal of Coastal Research, 213(213), pp.453–463. Available at: http://www.bioone.org/doi/abs/10.2112/03-0079.1 [Accessed May 15, 2012].
76
Rooker, J.R. et al., 1998. Postsettlement Patterns of Habitat Use by Sciaenid Fishes in Subtropical Seagrass Meadows. Estuaries, 21(2), p.318. Available at: http://link.springer.com/10.2307/1352478.
Scawthorn, C. et al., 2006. HAZUS-MH flood loss estimation methodology. II. Damage and loss assessment. Natural Hazards Review, 7, pp.72–81.
Scyphers, S.B. et al., 2011. Oyster reefs as natural breakwaters mitigate shoreline loss and facilitate fisheries. PloS one, 6(8), p.e22396.
Sekimoto, T. et al., 2013. A new criterion and probabilistic approach to the performance assessment of coastal facilities in relation to their adaptation to global climate change. Ocean Engineering, null(null). Available at: http://dx.doi.org/10.1016/j.oceaneng.2013.01.032 [Accessed June 4, 2013].
Sharp, R. et al., 2014. InVEST 3.0.0 User’s Guide, Stanford, CA. Available at: http://ncp-dev.stanford.edu/~dataportal/invest-releases/documentation/current_release/.
Shepard, C.C., Crain, C.M. & Beck, M.W., 2011. The protective role of coastal marshes: a systematic review and meta-analysis. PloS one, 6(11), p.e27374.
Siikamäki, J. et al., 2013. Blue Carbon: Coastal Ecosystems, Their Carbon Storage, and Potential for Reducing Emissions. Environment: Science and Policy for Sustainable Development, 55(6), pp.14–29.
Simmons, E.G. & Breuer, J.P., 1962. A study of red fish, Sciaenops ocelleta (Linnaeus ), and black drum, Pegonias cromis (Linnaeus). Publications of the Institute of Marine Science, University of Texas, 8, pp.184–211.
Stockdon, H.F. et al., 2012. National assessment of hurricane-induced coastal erosion hazards—Gulf of Mexico: U.S.G.S. Open-File Report 2012–1084, Reston, Va. Available at: http://pubs.usgs.gov/of/2012/1084/pdf/ofr2012-1084.pdf.
Stunz, G., Minello, T. & Rozas, L., 2010. Relative value of oyster reef as habitat for estuarine nekton in Galveston Bay, Texas. Marine Ecology Progress Series, 406, pp.147–159. Available at: http://www.int-res.com/abstracts/meps/v406/p147-159/.
Stunz, G.W., Minello, T.J. & Levin, P.S., 2002. Growth of newly settled red drum Sciaenops ocellatus in different estuarine habitat types. , 238(Fuiman 1994), pp.227–236.
Tagatz, M.E., 1969. Some relations of temperature accilmation and salinity to thermal tolerance of the blue crab, Callinectes sapidus. Transactions of the American Fisheries Society, 4, pp.713–716.
77
Tallis, H. et al., 2008. An ecosystem services framework to support both practical conservation and economic development. Proceedings of the National Academy of Sciences of the United States of America, 105, pp.9457–9464.
Texas Parks and Wildlife Department (TPWD), 2010. Oyster fishery management options, Austin, TX.
Texas Seagrant, 2013. The Risk of Rising Sea Level: Texas Universities Stand Ready and Able to Help Coastal Communities Adapt, College Station, TX.
The Bay Institute, 2013. Analysis of the costs and benefits of using tidal marsh restoration as a sea level rise adaptation strategy in san francisco bay, San Francisco, CA.
Thomas, J.L., Zimmerman, R.J. & Minello, T.J., 1990. Abundance Patterns of Juvenile Blue Crabs (Callinectes Sapidus) in Nursery Habitats of Two Texas Bays. Bulletin of Marine Science, 46(1), pp.115–125. Available at: http://www.ingentaconnect.com/content/umrsmas/bullmar/1990/00000046/00000001/art00010.
Thornton, E.B. & Guza, R.T., 1983. Transformation of wave height distribution. Journal of Geophysical Research, 88(C10), pp.5925–5938.
TPWD, 2002. Executive Summary: The Texas Shrimp Fishery. A report to the Governor and the 77th Legislature of Texas.,
Turner, R.E., 1977. Intertidal Vegetation and Commercial Yields of Penaeid Shrimp. Transactions of the American Fisheries Society, 106(5), pp.411–416.
Turner, R.E. & Brody, M.S., 1983. Habitat suitability index models: northern Gulf of Mexico brown shrimp and white shrimp. U.S. Dept. of Int. Fish Wildl. Serv. FWS/OBS-82/10.54., p.24.
U.S. Army Corps of Engineers (USACE), 2002. Coastal Engineering Manual (CEM), Report EM 1110-2-1100, Vicksburg, MS.
U.S. Army Corps of Engineers (USACE), 2003. Economic Guidance Memorandum 04-01; , Generic Depth-Damage Relationships for Residential Structures with Basements., Washington, D.C.
U.S. Army Corps of Engineers (USACE), 1963. Interim Survey Report, Morgan City, Louisiana and Vicinity, New Orleans, LA.
U.S. Army Corps of Engineers (USACE), 2010. Wave Information Studies Project Documentation, Vicksburg, MS.
78
U.S. EPA, 2013. The Social Cost of Carbon, Washington, D.C. Available at: http://www.epa.gov/climatechange/Downloads/EPAactivities/scc-fact-sheet.pdf.
U.S. Fish and Wildlife Service, 1997. Wetlands of the Galveston Bay system : anchoring the bay’s food chain, Webster, TX: Galveston Bay Foundation ;;U.S. Fish and Wildlife Service.
U.S. Fish and Wildlife Service (USFWS), 2009. National Wetland Inventory. Available at: http://www.fws.gov/wetlands/NWI/index.html [Accessed January 1, 2011].
United States Census Bureau (USCB), 2013. US Census County Based TIGER/Line; Texas, 2010 Census Block State-based Shapefile with Housing and Population Data, Washington, D.C . Available at: http://catalog.data.gov/dataset/tiger-line-shapefile-2010-2010-state-texas-2010-census-block-state-based-shapefile-with-housing.
United States Interagency Working Group on Social Costs of Carbon (U.S. IWGSCC), 2010. Technical Support Document: Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866, Washington, D.C. Available at: http://www.epa.gov/otaq/climate/regulations/scc-tsd.pdf.
Venkataramaiah, A., 1971. Study of the salinity relationship of the commercial shrimp (Family Penaeidae) in estuarine waters. Contract No. DACH39- 71-C-008 (Prog. Rep. June 1971-August 1971). U.S. Army.
Wallace, D.J., Anderson, J.B. & Rodriguez, A.B., 2009. Natural versus anthropogenic mechanisms of erosion along the upper Texas coast. Geological Society of America Special Papers, 460, pp.137–147.
Walsh, S. et al., 2013. Green and grey infrastructure: Valuation for corporate coastal hazard mitigation, Arlington, VA.
Wamsley, T. V et al., 2009. Influence of Marsh Restoration and Degradation on Storm Surge and Waves. , (July).
Wamsley, T. V et al., 2010. The potential of wetlands in reducing storm surge. Ocean Engineering, 37(1), pp.59–68.
Ward, G.H., 2012. THE BLUE CRAB: A SURVEY WITH APPLICATION TO SAN ANTONIO BAY, Available at: https://www.twdb.texas.gov/publications/reports/contracted_reports/doc/0900010973_BlueCrab.pdf.
Warner, N.N. & Tissot, P.E., 2012. Storm flooding sensitivity to sea level rise for Galveston Bay, Texas. Ocean Engineering, 44, pp.23–32.
Warren Pinnacle Consulting (WPC), 2011. Application of the Sea-Level Affecting Marshes Model (SLAMM 6) to Galveston Bay, Waitsfield VT.
79
Warren Pinnacle Consulting (WPC), 2009. Sea Level Affecting Marshes Model, Version 6; Technical Documentation, Waitsfield VT.
Yokel, B.J., 1966. A contribution to the biology and distribution of the red drum, Sciaenops ocellata. M.S. Thesis. University of Miami, Coral Gables, Fla., p.160.
Yoskowitz, D. et al., 2012. Assessment of Changing Ecosystem Services Provided by Marsh Habitat in the Galveston Bay Region, Corpus Christi, TX.
Yoskowitz, D.W., Gibeaut, J. & McKenzie, A., 2009. The Socio-Economic Impact of Sea Level Rise in the Galveston Bay Region - A report for Environmental Defense Fund, Corpus Christi, Texas.
Yu, J. et al., 2012. Soil organic carbon storage changes in coastal wetlands of the modern Yellow River Delta from 2000 to 2009. Biogeosciences, 9(6), pp.2325–2331.
Zein-Eldin, Z.P. & Renaud, M.L., 1986. Inshore Environmental Effects on Brown Shrimp , Penaeus aztecus , and White Shrimp , P . setiferus , Populations in Coastal Waters , Particularly of Texas. Marine Fisheries Review, 48(3), pp.9–19.
Zimmerman, R.J., Minello, T.J. & Rozas, L.P., 2002. Salt Marsh Linkages to Productivity of Penaeid Shrimps and Blue Crabs in the Northern Gulf of Mexico. In P. Weinstein & D. Kreeger, eds. Concepts and Controversies in Tidal Marsh Ecology. Kluwer, Dordreht and Boston: Springer Netherlands, pp. 293–314.
Zimmerman, R.J., T.J.M. & Zamora, G., 1984. Selection of vegetated habitat by brown shrimp, Penaeus aztecus, in a Galveston Bay saltmarsh. U.S. Fishery Bulletin, (82), pp.325–336.
Appendix A Land Class Classification for Storm Surge and Wave Modeling
In order to model storm surge and wave evolution in and around Galveston Bay, we
created a land cover map of the Galveston Bay region large enough to include all land classes
affected by a storm. We use this map to generate a database of Manning 𝑁 coefficients to model
flow resistance and total water level. We also used this map to gather information on the physical
characteristics of major vegetation classes present in the region.
We created a land cover map of the Bay region by merging the land cover map provided
by WPC (2011) with the Coastal Change Analysis Program (C-CAP) land cover data (NOAA
80
Coastal Services Center (NOAA-CSC) 2006). We re-classified those data based on the 2009
National Wetland Inventory (NWI) scheme, (U.S. Fish and Wildlife Service (USFWS) 2009) as
shown in Figure A-1.
Figure A-1: Merged land cover data (WPC (2011) and C-CAP dataset). The grey scale indicates the land cover data used as base conditions in SLAMM modeling and the red to blue color scale indicates the 2006 C-CAP land cover data that was merged with the SLAMM dataset for a more complete footprint.
Next, we associated each land class type with a corresponding Manning 𝑁 coefficient
(Table A-1). We generated this database from a compilation of various data published in the
literature (Mattocks & Forbes 2008; Bunya et al. 2010; Agnew 2012; Wamsley et al. 2009;
Wamsley et al. 2010). We ensured that estuarine marshes had a lower 𝑁value than salt marsh,
assuming that plant species in fresh marshes tend to be taller, more rigid, and more dense and,
therefore, more dissipative. We did not base our choice on model calibration to observed storms
(Wamsley et al. 2010).
81
Table A-1: Manning Coefficients Associated With Different Land Classes.
SLAMM Code SLAMM NAME Manning
N CCAP Code
Land Cover Type Manning N
1 Developed Dry Land 0.093 02
High Development 0.093
3 Swamp 0.074 03 Med Development 0.093
4 Cypress Swamp 0.074 04 Low Development 0.093
5 Inland Fresh Marsh 0.055 05
Open Development 0.03
6 Tidal Fresh Marsh 0.055 06 Cultivated 0.05
7 Scrub/Shrub 0.074 07 Pasture/Hay 0.05
8 Regularly Flood Marsh 0.045 08 Cropland 0.05
10 Estuarine Beach 0.03 09 Deciduous Forests 0.074
11 Tidal Flat 0.03 10 Evergreen Forests 0.074
12 Ocean Beach 0.03 11 Mixed Forest 0.074
15 Inland Open Water 0.02 12 Scrub/Shrub 0.074
16 Riverine Tidal 0.03 13 Palustrine Forest Wetland 0.074
17 Estuarine Water 0.02 14
Palustrine Scrub/Shrub Wetland 0.074
19 Open Ocean 0.02 15
Palustrine Emergent Wetland 0.055
20 Irregurlarly Flooded Marsh 0.055 16
Estuarine Forested Wetland 0.074
22 Inland Shore 0.03 17
Estuarine Scrub/Shrub Wetland 0.074
23 Tidal Swamp 0.074 18
Estuarine Emergent Wetland 0.045
19 Unconsolidated Shore 0.03
20 Bare Land 0.03
82
21 Water 0.02
Finally, we used the land cover map to gather information on the physical characteristics
of vegetation in and around the Bay. Because of the scarcity of information on physical
parameters for all vegetation types present in Galveston, we grouped different land class types as
described below and summarized in Table A-2:
• Swamp and forested wetland: cypress swamp, swamp, tidal swamp, palustrine forested
wetland, and estuarine forested wetland are grouped together and given the physical
characteristics of cypress, which is the dominant species in swamps and forested wetland
(Lester & Gonzalez, 2011).
• Scrub/shrub areas: scrub/shrub wetland, estuarine scrub/shrub wetland, palustrine
scrub/shrub wetland are characterized by woody vegetation less than 5m in height. While
this may contain a variety of species, including young trees, we assigned to all vegetation
types in that category the physical characteristics of Buttonbush.
• Forests: deciduous forests, evergreen forests, and mixed forests are grouped together and
considered upland forest.
• Low salt marsh: estuarine emergent wetland (CCAP) and regularly flooded marsh
(SLAMM) are grouped together and attributed the physical characteristics of Spartina
alterniflora.
• High salt marsh/brackish marsh: tidal fresh marshes and irregularly flooded marshes were
grouped together and represented by marsh hay cordgrass (Spartina patens).
• Fresh Marsh: palustrine emergent wetland (CCAP) and inland fresh marsh (SLAMM) are
grouped together and represented as giant cutgrass (Zizaniopsis miliacea), which is the
83
primary species occurring in fresh marshes (Lester & Gonzalez 2011)). Due to a lack of
data, we approximated the physical properties of this species by using data for big
cordgrass.
• Agriculture, Bare land, Open Water: land cover of these types are represented in the
model with their corresponding Manning coefficient (see Table A-1).
Table A-2: Physical Characteristics of Vegetation Associated with Different Land Classes
Land Cover Class Height [m]
Diameter [m]
Density [units/m2] Source
Swamps and Forested Wetlands 15 1.5 0.0172 Arborday.org
Scrub/Shrub 2.5 0.1016 0.0861 University of Florida, Alabama Forestry Department, and Davesgarden.com
Forests 10 0.2 0.0861 FEMA
Low Salt Marsh 0.8 6.4x10-3 172 FEMA guidance for the Chenier Plain Region
High Salt Marsh/Brackish Marsh
0.57 5.8x10-4 3,585 FEMA guidance for the Chenier Plain Region
Fresh Marsh 1.22 8.14x10-3 107 FEMA guidance for the Chenier Plain Region
84
Appendix B Marsh Acreage
To investigate the extent to which conversion of marshes to development has reduced
marsh coverage, we merged historic (1989 and 1979) NWI marshes with the current NWI
marshes (2004) to represent a possible high coverage future marsh footprint. Data from 1956
were not included because they are inadequate to distinguish marshes versus non-marshes. From
this merged dataset, we removed areas where the current LULC was classified as “Commercial”,
“Industrial”, “Residential”, “Governmental/Medical/Educational”, and “Roads” (Developed
Areas), according to the regional growth forecast produced by H-GAC (2013). That forecast
lacks coverage in Chambers County in GIS format. Therefore, we did not compute any loss of
marsh footprint in that area. This approximation is reasonable considering the low population
density in that region.
We find that, based on historical data, the marsh might have covered more than 2,162
km2 before present. This means that approximately 123.89 km2 of marshes may have been lost
due to direct conversion to development between 1979 and today (current marsh area is
approximately 2,038 km2). An additional 77 km2 may be lost due to future conversion to
development by 2040. Closer comparison of the land cover map in 2040 as predicted by H-GAC
and of the current marsh footprint shows that most of this loss will occur in small, isolated
patches. No cohesive and compact area of marsh is projected to be lost because of this forecasted
development.
As mentioned in Section 2.2, sea level rise is likely to be the main climate variable
expected to impact coastal wetlands in Galveston Bay. This would result in the conversion (and
sometimes net loss) of one wetland type to another (Warren Pinnacle Consulting (WPC) 2011).
In this section, we investigate the response of salt marshes to sea level rise.
85
Salt marsh zonation in Galveston Bay is tightly linked to tidal elevation, with the low
marsh dominated by Spartina alterniflora and the upper marsh by Spartina patens. Together
these habitats colonize available substrate from mean sea level to the upper edge of the saltwater
inundation zone (Feagin and Wu 2006). As sea levels rise, inundation frequency will change and
marshes will be required to reestablish proper zonation to maintain equilibrium in flood regimes.
In general, marsh that becomes permanently inundated as sea levels rise will likely die, and
marshes in areas with projections of high sea level rise combined with small tidal ranges are
most vulnerable (Craft 2009, Stevenson and Kearney 2009). Consequently, salt marshes in
Galveston Bay, which is a microtidal bay with a relatively high rate of SLR, are projected to
incur significant losses due to sea level rise (Warren Pinnacle Consulting (WPC) 2011).
Salt marshes may adapt to sea level rise through two mechanisms, vertical accretion of
the marsh platform and inland migration. Areas of Galveston Bay with ample sediment supply
may outpace sea level rise through accretion. However, in other areas, particularly West
Galveston where inorganic sediment is limiting (Ravens et al. 2009) and accretion rates fall well
below sea level rise rates (Ravens et al. 2009, Feagin et al. 2010), it might be harder for salt
marshes to adapt, or keep up. In that case, marsh will migrate inland, and adjacent non-salt marsh
areas will be converted to salt marsh. The potential for marsh to migrate inland depends on a
number of biophysical factors including slope, substrate, and the robustness of the marsh plant.
However, the biggest barrier to marsh migration is human development. Indeed, roads,
agricultural lands, building and other types of development are either protected from the impacts
of SLR in ways that exclude marsh migration, or cannot be converted to marsh habitat.
In this section, we use the InVEST Habitat Climate Adaptation model (beta version) to
predict where salt marsh migration may be blocked or facilitated, based on tidal elevation, slope,
86
accretion rate, erosion, sea level rise and land use/land cover information. This tool was
developed using the same principles as the SLAMM model, but focuses solely on salt marsh
areas. Consequently, it predicts future marsh footprints in regions located between mean sea
level and the ‘salt boundary’ (upper edge of tidal inundation) where the contiguous upper and
lower saltmarsh grow. This region encompasses the SLAMM categories of regularly and
irregularly flooded marsh.
First, we estimate the migration potential of salt marshes in Galveston Bay based on the
land class present landward of the marsh habitats. For each land class, we determine a migration
potential value, based on the likelihood that this land class can be converted to marsh habitat
(Table B1). Results of this analysis are shown in Figure B-1, where we show regions where
marsh is more likely than others.
Table B1: Marsh migration potential factors as a function of land class.
General Classification Migration Potential Examples of land class used
Developed Land 0.0 Low to high intensity development, multiple family dwelling units, etc…
Open Water 0.0 Inland and marine water Roads 0.0 Major highways, etc. Agricultural Land 0.3 Cultivated land, etc. Bare land 0.5 Bare land, grassland, open pasture, etc.
Forested Land 0.7 Evergreen forest, mixed forest, deciduous forest
Freshwater Wetland 1.0 Palustrine emergent wetland, palustrine forested wetland
Scrub/Shrub Wetland 1.0 Scrub/shrub wetland
Tidal Wetland 1.0 Estuarine emergent & brackish wetland, irregularly & regularly flooded marsh, etc.
Shoreline 1.0 Unconsolidated shoreline, beach
87
Figure B-1: Likelihood of marsh migration in Galveston Bay. Left: Outputs for land classes classified based on the likelihood that marshes will be permitted to migrate across different land classes. Right: Results where migration factors less than 1 are considered to be 0.
Next, we use the relationship between tidal elevation and marsh zonation to identify the
region that marshes may inhabit under future sea level conditions (Figure B-2). Salt marsh may
inhabit part or all of this region, depending on its vertical accretion rate. In addition to inland
migration, vertical accretion of the marsh platform through sedimentation or biological
accumulation of peat is a second adaptation strategy that may allow marshes to outcompete sea-
level rise. Here, we entered site-specific accretion rates to offset inundation from sea-level rise.
This represents a simplification over the SLAMM model, which incorporates the relationship
between tidal elevation and marsh productivity to calculate a more dynamic representation of
accretion rates. Results are shown in Figure B-2 and B-3.
88
Figure B-2: Potential zone of future marsh location. Left: future marsh zone corresponding to the region between MSL and the ‘salt boundary’ given 1m SLR in Galveston Bay. Right: future marsh zone in the Bay computed by taking into account vertical accretion.
Future marsh distribution
Figure B-3: Potential future marsh distribution in 2100 with (a) 1 meter SLR, (b) 1 meter SLR for a section of Galveston Island, (c) 2 meter SLR, and (d) the A1B mean (0.39 m) SLR scenarios. Marsh contained in diked areas (in pink) is assumed to remain unchanged until dikes are overtopped (>2m SLR). Marsh in red is blocked from migration by development, etc., and is assumed lost in 2100 (this marsh is automatically removed from the future marsh footprint). Marsh in varying shades of blue corresponds to ‘likelihood’ of marsh migration across the landscape based on land use (darker blue=increased likelihood). This depicts where marsh may be most imperiled by existing land use.
Validation with SLAMM model
89
Results from the InVEST tool were validated against those from the SLAMM model. We find
good alignment between SLAMM and InVEST predicted marsh distributions for the 1m, 2m and
A1B mean SLR scenarios (the latter two now shown) (Figure B-4). InVEST incorporates three
SLAMM habitat types (regularly flooded, irregularly flooded and transitional salt marsh) into
one habitat footprint (saltmarsh). There are some deviations between SLAMM and InVEST,
which may be accounted for by slight differences in the DEM or SLAMM’s more complex
accretion feedback loop. Furthermore, while marsh erosion by waves will be incorporated into
the InVEST tool, it was not included in this analysis.
a)
b)
Figure B-4: Validation of InVEST climate adaptation model against SLAMM results in East Bay (a) and in West Bay (b).
90
SLAMM Model Outputs
WPC (2011) computed changes in habitat and wetland coverage in Galveston Bay, for
different sea-level rise scenarios. Table B2 shows changes in land class types for the 1 m SLR
scenario.
Table B2: SLAMM Outputs Depicting Changes in Land Classes for 1 m Sea-Level Rise
Land cover Category SLAMM Code
2004 2050 2100 Area
[thousands of Ha]
Area [thousands
of Ha]
Change from 2004
Area [thousands
of Ha] Change
from 2004 Developed Dry Land 1 43.8 43.3 -1% 41.3 -6% Undeveloped Dry land 2 143.1 132.0 -8% 116.8 -18% Nontidal Swamp 3 9.2 9.3 1% 8.6 -6% Cypress swamp 4 0.0 0.0 -4% 0.0 -18% Inland fresh marsh 5 47.5 44.7 -6% 35.0 -26% Tidal fresh marsh 6 2.3 3.0 31% 1.7 -25% Transitional salt marsh/Shrub 7
7.6 10.1 33% 12.7 69% Regularly flooded salt marsh 8
8.7 13.2 52% 20.1 131% Estuarine beach 10 1.4 0.7 -52% 0.2 -82% Tidal flat 11 1.5 9.6 559% 16.5 1030% Ocean beach 12 0.5 0.4 -7% 0.7 52% Inland open water 15 10.5 8.6 -19% 8.1 -23% Riverine tidal open water 16
1.3 0.6 -57% 0.4 -73% Estuarine open water 17 162.0 170.5 5% 195.0 20% Open ocean 19 348.7 348.7 0% 348.9 0% Irregularly flooded salt marsh 20
24.3 18.6 -23% 8.4 -65% Inland shore 22 2.0 2.0 -2% 1.9 -6% Tidal swamp 23 2.4 1.6 -34% 0.4 -85%
91
Estuarine Wetlands 45.3 46.4 3% 43.4 -4% Freshwater Wetlands 56.7 54.0 -5% 43.7 -23% Open Water and Beaches 527.8 541.0 2% 571.6 8% Land 186.9 175.3 -6% 158.1 -15%
92
Appendix C Storm Surge
Coastal vegetation has the capacity to reduce storm surge by increasing the amount of
flow energy lost to bottom friction dissipation. Bottom friction over submerged lands is a
function of the Manning coefficient associated with different land classes. We estimate the
potential for wetlands to attenuate storm surge in West Bay by running the AdCirc storm surge
model (Luettich et al. 1992) for three storms conditions (Table C-1): a tropical storm (storm
“B0”), and a category 1 (“Z1”) and 2 (“Z2”) hurricane. All three storms had the same idealized
track (Figure C-1), and the storm parameters were held constant throughout the storm simulation.
Although these storms only impact West Bay and are idealized, they give an indication of the
surge reduction potential of wetlands in the Bay. We find that wetlands have the capacity to
reduce storm surge in the Bay, but that the amount of surge reduction is highly dependent on the
local topography.
Table C-1: Storm Parameters.
Simulation ID
Central Pressure [mb]
Maximum Wind Speed [kts]
Radius of Maximum Wind [nm] Storm Source*
B0 996 55 25 Eduardo 2008 Z1 982 75 25 Claudette 2003 Z2 970 85 25 Unnamed 1949
*The source indicates actual storms in the area from which central pressure and maximum wind speeds were obtained
93
Figure C-1: Storm track in Galveston Bay. The green dots indicate the eye path for the completed simulations. The hatched polygon and blue solid polygon represent the bands where the wind speeds are within 10 and 5% of the maximum offset from the storm path, respectively.
Figure C-2 shows the maximum water surface elevation generated Z2 storm induced
surge, in the absence of vegetation. Comparisons of surge elevation in the presence and absence
of wetlands in that region (Figure C-3) show that the wetlands have the potential to reduce surge
by as much as 20 cm. However, the role of wetlands, captured through the range of computed
attenuation values, is not uniformly distributed along the vegetated area. Wetlands are more
effective at reducing storm surge on the west side of West Bay than on its east side.
94
Figure C-2: Storm surge elevation in Galveston Bay generated by Storm Z2, in the absence of vegetation.
Figure C-3: Difference between maximum surge elevations in West Bay with and without habitat.
95
To compute the average rate of storm surge reduction by wetlands in West Bay, we
extracted profiles of the overland surge along the different transects shown in Figure C-3, for the
three storms modeled, B0, Z1 and Z2. Next, we calculated the slope of the difference between
surge computed in the presence and absence of habitats plots in regions where such difference
was approximately linear. However, because of the complex nature of the topography in that
region, most of the profiles analyzed produced different profiles that we could not confidently
use to compute a linear attenuation rate. Only a few transects (circled by a thick black oval in
Figure C-3) captured a clean profile from which an attenuation rate could be inferred. Figure C-4
shows an example of such profiles, for storms Z1 and Z2.
Figure C-4: Example profiles of storm surge in West Bay, in the presence and absence of wetlands. The value of maximum surge elevation associated with Z1 (green) and Z2 (black), with (solid) and without habitat (dashed). The maximum surge dips to 0 at topographic high points not inundated by the surge (top). The difference between with and without habitats for Z1 (green) and Z2 (black) (center). The topographic profile overlain with different colored segments representing the different habitats and corresponding Mannings N-values along the profile. The line representing forest is dashed because that is the only habitat not excluded in the without habitat simulation. Also, there is no forest present along this profile (bottom).
96
For the 5 transects used in our analysis, we estimated surge attenuation rates varying
from 19.8 cm/km of wetland (storm B0) to 15 cm/km (storm Z2). Our results are on the higher
side of the range of attenuation rates reported for other wetlands systems, but seem to be
reasonable (Table C-2). Outputs also show the difficulty in coming up with simple rules of
thumb and general guidelines about the role of wetlands at reducing storm surge.
Table C-2: Attenuation rates of storm surge by wetlands
Study Attenuation Rates [cm/km]
Forcing Manning N Method
Wamsley et al. (2010)
2.0, 3.3, 4, 5.6 960mb, 900mb (Cat 3, Cat 5?)
0.035 – 0.055 ADCIRC-
McGee et al. (2006)
4.0, 7.7, 10, 25 Hurricane Wilma (no direct hit)
NA Field Observations
USACE (1963) 1.7 – 20.0 NA NA NA Present Study 19.8, 28.5, 15.7 TS, Cat 1, Cat2 0.045 – 0.074 ADCIRC
Figure C-5: Rates of surge reduction as measured by different studies
97
Appendix D Description of Terrestrial and Blue Carbon Models’ Inputs
InVEST Terrestrial Carbon model
We used the InVEST terrestrial carbon model (version 2.5.6) to estimate the amount of
carbon stored at the initial time step in each land cover class within the project site boundaries.
Model inputs include the 2004 land cover map from SLAMM and estimates of the amount of
carbon in each land cover category. We included in our analysis the first three pools and
excluded the detrital layer because of a lack of data. Additionally, because local data are difficult
to obtain, we used data from other geographies in the northern Gulf of Mexico region (Figure
D-1) to determine the amount of organic carbon in each of these pools.
Figure D-1: Salt marsh soil carbon content values from published literature. The red horizontal line represents the value used herein. The blue and green horizontal lines represent maximum and minimum reported soil carbon values.
We assigned an average soil carbon content value of 222 Tons C/ha to the regularly
flooded salt marsh landcover class. This value is below the global average value of 390 Tons
98
C/ha reported by Chmura et al. (2003), whose database was dominated by soil carbon values
collected in tropical climates. We assigned soil carbon values to the remaining habitat types by
using conversion factors between the regularly flooded marsh category and the remaining
wetland habitats (Table C1). Consequently, for irregularly flooded and transitional marshes,
which are considered high and mid marsh, respectively, we assigned a conversion factor based
on the relationship between high, mid and low marsh carbon content found in Choi et al. (2001).
Additionally, because of a lack of data for the different swamp classifications in Texas, we
grouped all swamp landcover categories (non-tidal, cypress and tidal). We then computed a
conversion factor for this group based on the relationship between the global average for salt
marsh and the global average for all coastal wetlands values in Chmura et al. (2003). For fresh
marsh landcover types, we assumed that the carbon content in fresh marshes is approximately
1.21 times that found in salt marshes (Craft et al. 2009). For tidal flats, we calculated the
conversion factor using the relationship between the average carbon content of salt flats and
wetland soils found in Yu et al. (2012). Although tidal flats have very little plant biomass they do
store carbon more effectively than estuarine open water habitats. Finally, for estuarine beach and
ocean beach, we allocated an organic carbon content of 10 Tons C/ha based on data published by
Yu et al. (2012).
99
Table C1: Wetland carbon conversion factors
Land cover Category Conversion Factor
Soil Carbon [Tons/ha] Source
Regularly flooded salt marsh 1.00 222 Nontidal swamp 1.10 244 Chmura et al., 2003 Cypress swamp 1.10 244 Chmura et al., 2003 Inland fresh marsh 1.21 269 Loomis and Craft, 2010 Tidal fresh marsh 1.21 269 Loomis and Craft, 2010 Transitional salt marsh 0.52 115 Choi et al., 2001 Tidal flat 0.25 56 Yu et al., 2012 Irregularly flooded salt marsh 0.45 100 Choi et al., 2001 Tidal swamp 1.10 244 Chmura et al., 2003
Table C2: Elemental carbon in carbon pools
Land cover Category Carbon in Biomass [Tons/ha] Soil Carbon [Tons/ha] Above Ground Below Ground
Developed dry land 0 0 0 Undeveloped dry land 0 0 0 Nontidal swamp 20 10 244 Cypress swamp 42 24 244 Inland fresh marsh 20 10 269 Tidal fresh marsh 20 10 269 Transitional salt marsh 20 10 115 Regularly flooded salt marsh 20 10 222 Estuarine beach 0 0 10 Tidal flat 0 0 55 Ocean beach 0 0 10 Inland open water 0 0 0 Riverine tidal open water 0 0 0 Estuarine open water 0 0 0 Open ocean 0 0 0 Irregularly flooded salt marsh 20 10 100 Inland shore 0 0 0 Tidal swamp 20 10 244
Blue Carbon Model
The InVEST blue carbon model (beta version) computes carbon accumulation and decay
in soils over time. However, because this model was initially designed for salt marshes, the only
100
land cover classes that are modeled for accumulation and decay over time are the three marine
wetland or salt marsh categories: regularly flooded, irregularly flooded and transitional salt
marshes. Accumulation and decay that also occur in the fresh marsh and swamp land cover
classes were not incorporated into the analysis.
The accumulation rate used herein was set at 0.573 TonsC/ha/yr, which was converted
from the average rate of 210g CO2/m2/yr for tidal marshes in the coterminous United States
(Chmura et al., 2003). This rate is multiplied by the number of years between two time steps to
calculate the total amount of carbon sequestered at a given time horizon. Carbon stored in the
soil also decays through time. Here, we assigned a decay value of 0.75 TonsC/ha over 50 years
for all land cover types that convert to other land cover classes, except for when salt marsh
converts to tidal flats. For this latter type of transition, we assigned a decay value of 0.50. Since
the blue carbon model is not iterative and only calculates the accumulation and decay in one
step, these values represent a loss of 75% and 50%, respectively, of the total carbon in the
previous salt marsh class by the next time step.
From the above mentioned input layers, the model generates two maps, one depicting the
amount of accumulated carbon and the other the amount of carbon lost, at different time
horizons.
101
Appendix E Modeled Fisheries Species’ Habitat and Environmental Conditions
Preferences
In this appendix, we describe the life histories, commercial importance and the habitat
and environmental parameters of the fisheries species studied.
Blue Crab
The blue crab (Callinectes sapidus) commercial fishery is one of the most important
commercial fisheries in Texas with landings in 2010 of 3.4 million pounds, or 3.1 million dollars
(ex-vessel). Movements of blue crab within estuaries are directly related to life cycle stages,
season and environmental factors (Van Den Avyle, M.J. & Fowler.. 1984; Perry & McIlwain
1986). Female crabs migrate from estuary waters to coastal waters or inlets releasing their eggs,
which develop in nearshore Gulf waters through their later larval stages (Perry and McIlwain,
1986). Once they reach the megalopal stage they migrate back into estuarine waters settling as
juveniles primarily into marsh habitats, but also into seagrass meadows and other shallow
nearshore areas (Pardieck et al. 1999; Van Heukelem 1991; Ward 2012). Eventually, they vacate
their primary nursery habitats and spread into the upper reaches of the estuary before moving
into deeper water as adults (Ward 2012; Perry & McIlwain 1986).
Water depth does not influence juvenile distribution, and they can live in water
temperatures between 2 and 35ºC (Tagatz 1969), with salinities ranging from 0.0 to well over 30
ppt (Pardieck et al. 1999; Hammerschmidt 1982). However, they prefer soft, muddy sediments
and tend to prefer marsh and seagrass habitats to estuarine waters (Evink 1976; Perry & Stuck
1982).
Brown Shrimp
102
Brown shrimps (Farfantepenaeus aztecus) are also an important commercial fishery in
Texas and, together with other commercial shrimp species, form the largest fishery in the state.
Over 27 million pounds of shrimp caught in 2010, which is worth an ex-vessel price of over 91
million dollars (NMFS 2010). Brown shrimps occupy a wide range of habitat types and are
found in both marine and estuarine habitats along the Texas coast. They typically migrate to
offshore waters during spawning periods. After hatching, young larvae spend their early stages in
offshore marine waters and migrate into shallow waters of estuaries and coastal wetlands (Turner
& Brody 1983). After entering the estuary, they transform into juveniles where they settle into
seagrass meadows and marsh edge zones (Turner and Brody, 1983). As the juveniles grow into
adults they migrate into deeper estuarine waters, eventually returning to deeper offshore Gulf
waters to spawn.
Juvenile brown shrimps can survive fluctuations in salinity from 10 to 20 ppt (Gunter,
Christmas & Killebrew 1964; Zein-Eldin & Renaud 1986). They can also survive in waters as
cold as 2 ºC and to over 32ºC (Venkataramaiah 1971; Christmas, J.Y., Langley & Devender
1976; NOAA 1998). However, post-larval and juvenile shrimps prefer shallow (< 2m) regions
with mud or silt bottom substrates rich in organic matter. They can also be found in mixed, sandy
and shell shallow substrate types (NOAA 1998; TPWD 2002). Finally, juvenile brown shrimps
prefer seagrass meadows and salt and brackish marsh areas over other habitat types (Turner &
Brody 1983; NOAA 1998; Zimmerman, R.J. & Zamora 1984), with annual populations of brown
shrimp found to be directly correlated with temperature and the area of suitable marsh habitat
(Barrett & Ralph 1976).
Red Drum
103
Red drums (Sciaenops ocellatus) support an important recreational fishery in Texas –
commercial fishing for red drum in Texas was banned in 1981 to reduce fishing pressure.
Recreational catches are not well documented in Texas and actual landings cannot be verified;
however, the recreational fishery industry in Texas is said to contribute billions of dollars in
yearly revenue to the Texas economy (NOAA 2012).
Red drums are an estuarine-dependent species that utilize estuarine habitats primarily
during their larval and juvenile stages but can be found in deeper waters as adults (Buckley
1984). Adult red drums migrate from marine waters to areas adjacent to channels and passes for
spawning, where their eggs and larval stages can be transported into their estuarine nursery
habitats. After hatching, red drum larvae settle in both vegetated and un-vegetated shallow water
estuarine habitats and along marsh edges (Rooker et al. 1998). Once they reach the juvenile stage
they stay in these protected areas of the estuaries, preferring grassy clumps or muddy bottoms
(Simmons & Breuer 1962). Older juveniles are found along shallow marsh edges or in marshes
until they mature when they start spending more time in marine waters (Yokel 1966).
Juvenile red drum tolerate a wide range of salinities and temperatures. They have been
found in salinities ranging from 0 to 50 ppt and temperatures ranging from 13 to 28 ºC (Perret,
W.S., Weavers, J.E., Williams, R.O., Johansen, P.L., McIlwain, T.D., Raulerson & Tatum 1980;
NOAA 1998). Juveniles tend to be found in waters swallower than 2 meters (Buckley, 1984) in a
wide variety of bottom sediment types including mud, silt, clay and sand.
In Texas larval and juvenile stages of red drum have been found to be associated with
seagrass meadows and marsh edge habitat (Yokel, 1966). Studies have also indicated that marsh
104
and seagrass habitat restoration can increase survival and growth rate of red drums (Levin &
Stunz 2005; Stunz et al. 2002).
Southern Flounder
Southern flounder (Paralichthys lethostigma) is another important fishery species in
Texas, both recreationally and commercially; however its economic impact is much less than the
blue crab and brown shrimp. Commercial landings of unclassified flounder species were around
26,000 pounds in 2010, worth an ex-vessel value of around $62,000 (NMFS 2010, commercial
landings).
Southern flounders recruit to estuaries early in their life cycle and are the dominant type
of flounder in the Gulf of Mexico that use estuary nursery grounds (McEachran & Fechhelm
2006). Adult southern flounder are commonly found along the shores, bays and lagoon of coastal
Texas. They become gravid and move offshore into the gulf during the fall and early winter to
spawn, although some stay within the estuary all year without becoming gravid (Ginsburg 1952).
Once released in the Gulf the eggs and resulting larval stages remain in marine waters for less
than two months before migrating into the estuaries (Arnold et al., 1977) where juvenile southern
flounder settle in shallow estuarine waters.
Juvenile southern flounder can withstand salinities ranging from 0.5 to 30 ppt. They have
been observed in water temperatures ranging from 5 to over 30 ºC. Juveniles prefer mud or silt
substrates but also can be found occasionally on sandy substrates (Enge & Mulholland 1985;
Packer et al. 1999). They are known to settle in marsh and seagrass habitats but are also found in
non-vegetative areas close to the water marsh interface (Glass et al. 2008). Because of this
105
association with marsh and seagrass habitats juvenile southern flounder in Texas are generally
found in waters shallower than 2 meters (Gutherz 1967; Glass et al. 2008; Packer et al. 1999).
106
Appendix F Fishery Habitat Change
We assess changes in the location and quantity of land and marine habitats to estimate
current and future amounts of preferred habitat for four different commercially and recreationally
important fish species in Galveston Bay: (1) Blue Crab (Callinectes sapidus), (2) Brown Shrimp
(Farfantepenaeus aztecus), (3) Southern Flounder (Paralichthys lethostigma), and (4) Red Drum
(Sciaenops ocellatus). Since these species are present in the Bay’s wetlands primarily during
their larval and juvenile stages, we only consider these life stages in our analysis (Appendix E).
Figure F-1 shows the modeling framework adopted herein. The most important and
common habitat and environmental parameters that define where these larval and juvenile fish
can live are sediment type, minimum and maximum depth, temperature, salinity and turbidity.
To apply the InVEST fisheries habitat model in Galveston Bay, we collected information about
the environmental and habitat preference ranges for each species of interest (Appendix E).
However, because of the poor resolution of publicly available data, we did not include sea
surface temperature, salinity and turbidity in our analysis.
From bathymetric and topographic, sediment and living habitat location information, the
model first extracts the maximum livable area for a given species’ life-stage based on depth
preferences. Those are waters between -36 m and MHW for blue crab, and between -2 meters
and MHW for the three remaining species. Next, the model extracts regions within the defined
area that have the types of sediment preferred by that species (Table F-1). However, because the
sediment dataset that we used is static and does not contain any future projection of sediment
types in the Bay, it was only used to delineate the maximum offshore extent of habitat suitable
for the different fisheries listed previously; sediment preference was not used to compute the
amount of suitable habitats in the future. Then the model incorporates information about location
107
and quantity of available living habitats for those species by generating habitatat suitability maps
based on the species preferences (Table F-2).
We obtained changes in the areal extent of estuarine and wetlands areas from SLAMM
outputs (Warren Pinnacle Consulting (WPC) 2011). Lastly, using information from the different
bathymetric and environmental layers, the model computes the amount of suitable habitat for that
species. We conducted this process for the different time horizons considered, based on updated
land cover maps.
Figure F-1: Fishery Habitat model framework
Table F-1: Surficial sediment preferences by species used for the spatial fishery modeling in Galveston Bay, Texas.
Common Name Sediment Classes and Species Preferences
Clayey Silt Gravel Gravelly
Sediment Sand Sand, Clay, Silt
Sandy Silt Silt Silty
Clay Brown Shrimp x x x Red Drum x x x x x x Southern Flounder x x x x x x Blue Crab x x x x
108
Table F-2: SLAMM habitat preferences by species used as parameters in the spatial fishery modeling in Galveston Bay, Texas.
Land cover Category SLAMM Code
Species Habitat Preference Blue Crab
Brown Shrimp
Red Drum
Southern Flounder
Swamp 3 Cypress swamp 4 Inland fresh marsh 5 Tidal fresh marsh 6 x Transitional salt marsh 7 x x x x Regularly flooded salt marsh 8 x x x x Estuarine beach 10 Tidal flat 11 x x Ocean beach 12 Ocean flat 13 x Inland open water 15 Riverine tidal 16 x x Estuarine water 17 x x Tidal creek 18 x x x Open ocean 19 Irregularly flooded salt marsh 20 x x x x Inland shore 22 Tidal swamp 23 Vegetated tidal flat 25 x x x x Back shore 26
Results from the model provide a first estimate of the changes in habitat for each key
species under different sea-level rise scenarios (Figure F-2, Figure F-3, Figure F-4). Under a 1 m
SLR scenario, the total amount of habitat, and marsh habitat in particular, increases for all
fisheries (Figure F-5), with the greatest increases occurring between 2050 and 2100. If wetlands
are allowed to migrate, riverine tidal areas are the only habitat that will shrink; they will lose
slightly more than two-thirds (-67%) of their surface area by 2100. The area of tidal fresh marsh
and regularly flooded marsh will more than double in 2100, and the amount of irregularly
flooded and transitional salt marsh will increase by 31 and 95%, respectively. In addition, blue
109
crabs will benefit from a modest increase in estuarine area (+5% by 2050 and +21% by 2100),
while red drums will see modest changes (-4% by 2050 and +6% by 2100).
110
Figure F-2: Habitat for juvenile blue crab in 2004 and 2100 under a 1 meter of SLR scenario.
111
Figure F-3: Habitat for juvenile southern flounder and brown shrimp in 2004 and 2100 under a 1 meter of SLR scenario.
112
Figure F-4: Habitat for juvenile red drum in 2004 and 2100 under a 1 meter of SLR scenario.
113
Figure F-5: Change in habitat area between 2004 and 2050 and 2004 and 2100 for a 1 m SLR scenario, assuming that wetlands can migrate (left) or cannot migrate (right).
Overall, all fisheries will experience net gains in habitats if wetlands are allowed to
migrate. Blue crabs will experience a net increase of 13% in total habitat area by 2050 and 44%
by 2100. More importantly, they will experience a 37% gain of wetland habitat, or “preferred
habitat”, by 2050 and a 117% gain by 2100 (Figure F-6). Southern flounder and brown shrimp,
which depend mostly on wetlands, will also benefit from a 37% in habitat area by 2050 and
117% increase by 2100 (Figure F-7). Finally, red drums will benefit from a 16% increase in total
habitat area and 37% increase in marsh habitat by 2050 and a 57% increase in total habitat and
an 11% increase in marsh habitat by 2100 (Figure F-8). The gains in marsh habitat are greater
than gains in total habitat for those fisheries because the amount of open water area decreases.
114
Figure F-6: Gains in habitat by type for juvenile blue crab in Galveston Bay, Texas under a 1 meter of SLR scenario.
Figure F-7: Gains and Losses in habitat by type for juvenile southern flounder and brown shrimp in Galveston Bay, Texas under a 1 meter of SLR scenario.
Figure F-8: Gains and Losses in habitat by type for juvenile red drum in Galveston Bay, Texas under a 1 meter of SLR scenario.
115
If wetlands are not allowed to migrate, the key species we examined will still experience
the same net gains in total habitat area, but those gains will mostly come from additional gains in
estuarine and open water areas because they will experience losses or little gain in marsh areas.
In particular, blue crabs will experience a 17% decrease in wetland habitat by 2050 and a modest
17% gain by 2100 (Figure F-6). Southern flounder, brown shrimp and red drum will see almost
no change in marsh habitat by 2050 (-1%) and a modest 17% gain by 2100 (Figure F-7 and
Figure F-8).
Consequently, most fisheries are likely to experience positive gains in habitat as sea level
rises. More importantly, they are likely to experience net gains in wetland areas by 2050 and
even greater gains by 2100, if marshes are allowed to migrate. If marshes are not allowed to
migrate, fisheries will mostly benefit from even wider open areas. They will mostly experience
little positive change or small losses in the amount of marsh habitat that they can utilize.
116
Appendix G Marsh Habitat Captured in the Fishery Habitat Model
In the Fishery Habitat model, we determine the change in salt marsh habitats area
(regularly flooded, irregularly flooded and transitional) with sea level rise. To do so, we capture
the total area of each of the three salt marsh categories withindifferent elevation ranges. Since
marsh habitat is the preferred habitat for most of the important fisheries in the Bay (see Section
3.2.5), the ability of those masks to capture salt marsh habitat is important.
We created two masks, for the different species studied herein. For Blue Crab, we created
a mask, referred to as the max mask, that extends from -36 m to MHW. For all other species, the
mask, referred to as the min mask, extends from -2 m to MHW. In this section, we compare the
relative effectiveness of both masks.
First, comparison of the area of the two different masks (Figure G-1) shows that the -36
mask (max mask) is more than five times larger than the -2 m mask (min mask). The area of the -
36 m mask increases under each time step, for all SLR scenarios. While the -2 meter DEM mask
area also increases under each time step, overall, it decreases in size by 13% under the two meter
SLR scenario during the final period between 2050 and 2100 This indicates that this mask best
captures the rapid SLR experienced under the 2 m scenario, with the nearshore areas becomes
deeper as low lying, flatter areas become inundated.
117
Figure G-1: Area comparison between the -36 meter (max) and the -2 meter (min) DEM masks under 1 meter, 2 meter and IPCC mean SLR scenarios
The spatial extent of salt marsh habitat that is captured in the DEM mask under the 1 m
SLR scenario are shown for the initial conditions, 2050 and 2100 (Figure G-3, Figure G-2, and
Figure G-4). We find that only a portion of the marsh habitat falls within the footprints of the
DEM mask for the initial time period (Figure G-3). However, for the 2050 and 2100 time
horizons, the proportion of salt marsh habitat that lies outside the mask decreases.
118
Figure G-2: Salt marsh habitats in 2004 that fall inside and outside of the -2 meter to MHW DEM spatial extent.
Figure G-3: Salt marsh habitats in 2050 that fall inside and outside of the -2 meter to MHW DEM spatial extent under a 1 meter of SLR by 2100 scenario.
119
Figure G-4: Salt marsh habitats in 2100 that fall inside and outside of the -2 meter to MHW DEM spatial extent under a 1 meter of SLR by 2100 scenario.
The full results for the DEM marsh habitat analysis are shown in Figure G-5. Overall,
under the initial 2004 conditions, less than 36% of the total area of the three SLAMM salt marsh
landcover types fall within the DEM mask. Irregularly flooded marsh, which occupies the
highest elevations of the three marsh categories, has 26% of its total area within the bounds of
the DEM mask, but its initial area is also much larger than the other salt marsh types. Fifty-five
percent (55%) of the regularly flooded marsh, which lies lower in elevation, is captured, while
71% of transitional salt marsh area is captured.
As sea level rises, more of the marsh habitat is captured by the mask. For the 2m SLR
scenario, over 91% of all salt marsh area in Galveston Bay is covered by the mask in 2100. For
the 1 m SLR scenario, this number is reduced to 83%. If we assume that marshes are not allowed
to migrate (no new marsh created as sea level rises), the final percentages of salt marsh covered
by the DEM mask are very similar to the marsh migration results with 87%, 82% and 41% under
120
the 2 m, 1 m and IPCC mean SLR scenarios, respectively. Finally, comparison of salt marsh area
covered with the -36 m and -2 m mask shows that under the 1 m and IPCC mean SLR scenarios,
the percent of marsh that falls within the two different DEM masks are the same. Under the 2 m
SLR scenarios the -2 meter DEM mask contained 88% (migration) and 41% (no migration) of
the total salt marsh area compared to 91% and 45% under the -36 meter mask.
In conclusion, although both masks do not capture all the habitats present in the Bay, they
are effective in capturing a large majority of the lower salt marsh boundaries found within the
study area.
Figure G-5: Salt marsh habitats that fall within the DEM masks for the -36 meter to MHW (A) and the -2 meters to MHW (B) depth ranges under six SLR and marsh migration scenarios for regularly flooded (RF), irregularly flooded (IF) and transitional salt marsh (TSM).
121